This action might not be possible to undo. Are you sure you want to continue?

Department of the Army

**U.S. Army Corps of Engineers
**

Washington, DC 20314-1000

EM 1110-21415 5 March 1993

Engineering and Design HYDROLOGIC FREQUENCY ANALYSIS

**Distribution Restriction Statement
**

Approved for public release; distribution is unlimited.

DEPARTMENT OF THE ARMY U.S. Army Corps of Engineers Washington. D.C. 20314-1000

EM 1110-2-1415

CECW-EH-

1’

Engineer Manual No. IllO-Z-1415 Engineering and Design HYDROLOGIC FREQUENCY ANALYSIS 1. m. This manual provides guidance and procedures for frequency flows, precipitation. water surface elevation, and flood damage. Appk&&& This manual applies to major subordinate responsibility for the design of civil works projects.

9 -. . .

5 March 1993

analysis of: flood flows, low

commands,

districts. and laboratories

having

3. u. Frequency estimates of hydrologic. climatic and economic data are required for the planning, design and evaluation of flood control and navigation projects. The text illustrates many of the statistical techniques appropriate for hydrologic problems by example. The basic theory is usually not provided, but references are provided for those who wish to research the techniques in more detail. FOR THE COMMANDER:

WILLIAM D. BROk’N Colonel, Corps of Engineers Chief of Staff

DEPARTMENT OF THE ARMY U.S. Army Corps of Engineers Washington, D.C. 20314-1000 CECW-EH-Y Engineer Manual No. 1110-2-1415 Engineering and Design HYDROLOGIC FREQUENCY ANALYSIS Table of Contents Subject CHAPTER 1 INTRODUCTION Purpose and Scope .................... References .......................... Definitions .......................... Need for Hydrologic Frequency Estimates ... ......... Need for Professional Judgement FREQUENCY ANALYSIS Definition ........................... Duration Curves ...................... Selection of Data for Frequency Analysis Graphical Frequency Analysis ............ Analytical Frequency Analysis ........... FLOOD FREQUENCY ANALYSIS Introduction ......................... ........ Log-Pearson Type III Distribution Weighted Skew Coefficient .............. Expected Probability ................... Risk .............................. ....... Conditional Probability Adjustment Two-Station Comparison ................ Flood Volumes ....................... Effects of Flood Control Works on Flood Frequencies .................... ................ Effects of Urbanization LOW-FLOW FREQUENCY ANALYSIS Uses .............................. Interpretation ........................ Application Problems ..................

EM 1110-2-1415

5 March 1993

Paragraph

Page

l-l 1-2 l-3 1-4 1-5 2-1 2-2 2-3 2-4 2-5 3-l 3-2 3-3 3-4 3-5 3-6 3-7 3-8 3-9 3-10 4-l 4-2 4-3

I-1 I-1 1-l 1-l l-2 2-l 2-3 2-6 2-10 2-12 3-l 3-l 3-6 3-7 3-10 3-11 3-13 3-18 3-22 3-28 4-l 4-1 4-l

CHAPTER

2

...

CHAPTER

3

CHAPTER

4

i

. ... Procedure ..... ....... ..... .. .. .. Regional Frequency Analysis .. ..... . ... ..... STATISTICAL RELIABILITY CRITERIA Objective . . .. . ... Practical Guidelines ... .. . ... .. . ... . Cautions .... Regression by Graphical Techniques . ..... .. . ..... Paragraph Page 5-I 5-2 5-3 6-l 6-2 6-3 6-4 7-l 7-2 7-3 8-1 8-2 8-3 TO REGIONAL 9-l 9-2 9-3 9-4 9-5 9-6 9-7 9-8 9-9 9-10 9-11 10-l 10-2 10-3 11-l 11-2 5-l 5-2 5-2 6-1 6-l 6-2 6-2 7-1 7-l 7-3 8-l 8-l 8-1 CHAPTER 6 CHAPTER 7 DAMAGE-FREQUENCY RELATIONSHIPS Introduction ........ . ... ... . . . Factors Responsible of Nondetermination Multiple Linear Regression Example .... . ANALYSIS . .. from Precipitation STAGE(ELEVATION)-FREQUENCY Uses .... . . . Computation of Expected Annual Damage ........ . ....... ... ...... .... . ...... ........ . REGRESSION ANALYSIS AND APPLICATION STUDIES Nature and Application . ... Partial Correlation .. . . ..... .. ..... . . . .. .. ... . . ....... ....... .... .... . .. ...... .. . The CorrelationCoefficient and Standard Error ... ... . ... .. . .. . .... . .. . .. Verification of Regression Results ...... . ... . . . .. ... ....... . . .... Available Regional Information Derivation of Flood-Frequency Relations ...... .. .. . . . . . .. ... .... ... . Expected Probability . ..... . .EM 1110-2-1415 5 Mar 93 Subject CHAPTER 5 PRECIPITATION FREQUENCY ANALYSlS General Procedures ... Equivalent Annual Damage ...... . .. .... . .... ....... Calculation of Regression Equations .. . Reliability of Frequency Statistics .... .. ... ............ FREQUENCY OF COINCIDENT FLOWS Introduction . .. ... . ANALYSIS OF MIXED POPULATIONS Definition . .. . . ..... .. .. . .... ....... .. A Procedure for Coincident Frequency Analysis CHAPTER 8 CHAPTER 9 9-l 9-l 9-3 9-4 9-7 9-8 9-8 9-10 9-10 9-10 9-11 10-l 10-l 10-2 11-l 11-l CHAPTER 10 CHAPTER 11 ii . .... . ... ...... . ... Simple L’ink’Regression Example ... Reliability of Frequency Curves . .. .. . Frequency Distribution . Stage Data . .. ..... ............ .. .. ... .. ..

... Deviates for the Expected Probability Adjustment .... GLOSSARY ............... .......................... Median Plotting Positions .... Confidence Limit Deviates for Normal Distribution ............ Deviates for Pearson Type III Distribution Normal Distribution ....... Values of Chi-Square Distribution Values of the F Distribution ............. Monthly Streamflow Model ............. Binomial Risk Tables ........ Applications ........ EXAMPLES OF RELIABILITY TESTS FOR THE MEAN AND STANDARD DEVIATION ............... Data Fill In .............. Outlier Test K Values (10 Percent Significance Level) ......................EM 1110-2-1415 5 Mar 93 Subject CHAPTER I2 STOCHASTIC HYDROLOGY Introduction .................. Daily Streamflow Model ......................... Reliability ..... SELECTED BIBLIOGRAPHY References and Textbooks ........ ............ STATISTICAL TABLES ................... Paragraph Page 12-l 12-2 12-3 12-4 12-5 12-6 12-7 12-8 A B 12-l 12-l 12-I 12-3 12-6 12-6 12-6 12-7 A-l A-4 APPENDIX A APPENDIX APPENDIX APPENDIX APPENDIX APPENDIX B C D E F B-l C-l D-l E-l F-l F-2 F-4 F-6 F-7 F-8 F-9 F-10 F-11 F-12 F-16 F-17 F-18 COMPUTATION PROCEDURE FOR EXTREME VALUE (GUMBEL) DISTRIBUTION .............. Mean-Square Error of Station Skew Coefficient ..... ........ Percentages for the Expected Probability Adjustment ........................................................... Application In Areas of Limited Data ....... HISTORIC DATA ................................ Basic Procedure ......................................................................... 111 ............. Computer Programs ........ Percentage Points of the One-Tailed t-Distribution ..............

. Example of Graphical Frequency Analysis .................. Daily Reservoir Elevation-Duration Curve ............ Probability Paper .. Frequency Curve.. Number 2-la 2-lb 2-2 2-3 2-4 2-5 2-6a 2-6b 3-l 3-2 3-3 3-4 3-5 3-6 3-7 3-8a 3-8b 3-9 3-10 3-11 3-12 3-13 4-l 5-1 6-1 6-2 6-3 6-4 7-l 8-l 9-l 9-2 9-3 9-4 9-5 9-6 9-7 10-I 11-l 12-la l2-lb Page 2-2 2-2 2-3 2-5 2-9 2-12 2-14 2... Coordination of Flood-Volume Statistics ... Annual Peaks and Sequential Computed Skew by Year ............... Partial Duration Frequency Curve....... Flood-Volume Frequency Relations ..... Schematic for Computation of Expected Annual Damage ... .... Low-Flow Frequency Curves .................................... Maximum Reservoir Elevation-Frequency Curve .... Storage Requirement Determination ........... Data Estimation from Regression Line ................ Log-Log Paper ..................... Annual Precipitation ............................. Regression Analysis for Regional Frequency Computations Regional Analysis Computations for Mapping Errors ............................... Flow-Frequency Curve.... Seasonal Variation of Elevation-Duration Relations ...... Illustration of Chronologic Sequence and Arrayed Flood Peaks ........................... Confidence Limit Curves based on the Non-central t-Distribution Cumulative Probability Distribution of Exceedances per 100 Years Two-Station Comparison Computations .......... Daily Flow-Duration Curves for Each Month .... Hurricane and Combined Flood Frequency Curves Illustration of Water-Surface Profiles in Coincident Frequency Analysis ........................ Derived Stage-Frequency Curves.... Data Estimation with Addition of Random Errors ... Observed and Two-Station Comparison Frequency Curves ....... Example With-Project versus Without-Project Peak Flow Relations Example Without-Project and With-Project Frequency Curves . Example Multiple Linear Regression Analysis ...EM 1110-2-1415 5 Mar 93 List of Figures Title Histogram and Probability Density Function ............................................... Typical Effect of Urbanization on Flood Frequency Curves . Computation of Simple Linear Regression Coefficients Illustration of Simple Regression ................................... Frequency Curve with Confidence Limit Curves .... Partial Duration Frequency Curve............ Regional Map of Regression Errors ............. Annual Frequency Curve ..... Daily Flow-Duration Curve ............. ....................... Unregulated and Regulated Conditions ....................... Sample and Theoretical Cumulative Distribution Functions ..........14 3-4 3-8 3-9 3-16 3-17 3-20 3-21 3-23 3-23 3-25 3-25 3-27 3-27 3-29 4-3 5-l 6-1 6-3 6-3 6-4 7-2 8-2 9-5 9-6 9-9 9-14 9-15 9-15 9-18 10-3 11-2 12-2 12-2 iv ..... Unregulated and Regulated Conditions Rating Curve for Present Conditions .... Nonhurricane......... Flood-Volume Frequency Curves .......

Number 2-1 2-2 2-3 3-l 3-2 4-l Page 2-4 2-11 2-13 3-3 3-18 4-2 .. High-Flow Volume-Duration Data ....... Low-Flow Volume-Duration Data ......... Sequential and Arrayed with Plotting Positions Partial Duration Peaks... Computed Frequency Curve and Statistics ....EM 1110-2-1415 5 Mar 93 List of Tables Title Daily Flow-Duration Data and Interpolated Values . Arrayed with Plotting Positions .............. V ........................ Annual Peaks...............

etc.. Need for Hvdrolonic Freauencv Estimates. This manual provides guidance in applying statistical principles to the analysis of hydrologic data for Corps of Engineers Civil Works activities. References cited in the text and a selected bibliography of literature pertaining to frequency analysis techniques are contained in Appendix A. Two basic problems exist for most hydrologic applications. zoning. l-l . some types of data are noted which usually do not fit any theoretical distributions. Frequency analysis should not be done without consideration of the primary application of the results. and “Hydrologic Frequency Analysis” (41). many of the statistical techniques appropriate for hydrologic problems. climatic and economic data are required for the planning. a single theoretical frequency distribution does not always fit a particular data-type equally well in all applications. Frequency estimates of hydrologic. a. pumping plants. These plans may consist of combinations of structural measures such as reservoirs. water use priorities. Also. precipitation amounts. and nonstructural measures such as flood proofing. General Guidance. levees. ADDlications. Objective. The text illustrates. hydroelectric power plants. design and evaluation of water management plans. river stages. by statistical standards. sediment loads. The probability estimates from these data are used in evaluating the economic.EM 1110-2-1415 5 Mar 93 ENGINEERING AND DESIGN HYDROLOGIC FREQUENCY ANALYSIS CHAPTER 1 INTRODUCTION I. ’ Numbered references refer to Appendix A. but references are provided for those who wish to research the techniques in more detail. b. The basic theory is usually not provided. Purpose and SCOD~. etc. Definitions. This manual provides guidance in fitting frequency distributions and construction of confidence limits. And secondly. l-3. number of stations to be included. insurance programs. l-4. resulting in uncertainty as to the true probability. The objective of frequency analysis in a hydrologic context is to infer the probability that various size events will be exceeded or not exceeded from a given sample of recorded events. l-2. The techniques described herein are taken principally from “Guidelines for Determining Flood Flow Frequency” (46)‘. The application will have a bearing on the type of analysis (annual series or partial duration series). lake stages. First the sample is usually small. flood damage. Techniques are presented which can possibly reduce the errors caused by small sample sizes. c. Appendix B contains a list of definitions of terms common to hydrologic frequency analysis and symbols used in this manual. The data to be analyzed could be streamflows. water demands. References. by example. “Statistical Methods in Hydrology” (l). etc. storm surge levels. channels. Selected Bibliography.1. social and environmental effects of the proposed management action.

EM 11 IO-2 1415 5 Mar 93 whether regulated frequency curves will be needed. etc. There may be applications where more complex joint or conditional frequency methods. Statistical analyses alone will not resolve all frequency problems. Need for Professional Judnment. l-5. It is not possible to define a set of procedures that can be rigidly applied to each frequency determination. that were considered beyond the scope of this guidance. l-2 . The user of these techniques must insure proper application and interpretation has been made. A frequency study should be well coordinated with the hydrologist. the planner and the economist. will be required. The judgment of a professional experienced in hydrologic analysis should always be used in concert with these techniques.

This curve is termed the cumulative distribution function. The accumulated area then represents non-exceedance probability or percentage (Figure 2. The probability of a certain magnitude event recurring again in the future. A pictorial display of the frequencies within each class interval is called a histogram (also known as a frequency polygon). Freaue cy of the statistical techniques that are applied to hydrologic data (to enable infer&es?% made about particular attributes of the data) can be labeled with the term “frequency analysis” techniques.1b). Definition. the area is accumulated from the smallest event to the largest.1. the enveloping line of the frequency distribution will approach a smooth curve. b. it is necessary to establish class intervals (arbitrary subdivisions of the range) and define the frequency as the number of events that occur within a class interval. Another means of representing the frequency is to compute the relative frequency. e. To have a perspective of the importance of the count. The term “frequency” usually connotes a count (number) of events of a certain magnitude.total number of events A graph of the relative frequency values is called a frequency distribution or histogram. Area accumulated in this manner represents exceedance probability or percentage. It is more common in hydrologic studies to accumulate the area from the largest event to the smallest. Figure 2-la. w In hydrologic studies.. Presentation of the data in this form is accomplished by accumulating the probability (area) under the probability density function. if the variable describing the events is continuous. 2-l = ni/N (2-l) 1 . where: f i = relative frequency of events in class interval i ni = number of events in interval i N . Sometimes the number of events within a specified time is used to give meaning to the count. This curve is termed the probability density function (Figure 2-la). (as are most hydrologic variables). the total number of events (sample size) must also be known. As the number of observations approaches infinity and the class interval size approaches zero.EM 1110-2-1415 5 Mar 93 CHAPTER 2 FREQUENCY ANALYSIS 2. Relative Freauency . is near zero. In most statistical texts. two daily flows were this low in 43 years. the probability of some magnitude c being’exceeded (or no: exceeded) is usually the primary interest.g. Therefore. a. The relative frequency is simply the number of events in the class interval divided by the total number of events: f.

Sample and Theoretical Cumulative 2-2 Distribution Functions. MD 5 2 b b = a B e a 0. Potonmc River 8t Point of Rock. *. ..a Peaka 6.la..EM 1110-2 1415 5 Mar 93 Potomac Riror at Point of Rocka.a 8.8 0 Logarithm of Annud Figure 2. ia ” ‘. 1 0 1 .... 1. : ” 1 :. Histogram and Probability Density Function.6 0 Logarithm of Annual Peakm Figure 2.8 .4 . 1. MD J I p.4 4.I b. 0 4 1. 0.

For this reason. . One of the first steps in preparation of a duration curve is dividing the range of the data into class intervals. The daily flow-duration curve cannot be considered a frequency curve in the true sense. The class intervals in Table 2-1 are based on a logarithmic distribution of the flows.8 99 se Percent of 68 Time Excmmdrd 10 1 . The events for flow-duration curves are usually mean daily flow values. the relative frequency curve is skewed to the right. The computation of flow-duration curves was probably the first attempt to analyze hydrologic data by statistical techniques.e1 Figure 2-2. because the daily flow on a particular day is highly correlated with the flow on the preceding day. provides a plot that is easily read at the extremities of the data. the abscissa is labeled “percent of time. Daily Flow-Duration 2-3 Curve.” 99.EM 1110-2-1415 5 Mar 93 2-2. Comoutation.99 99. Therefore. The flows tend to be grouped near the low end with very few large flows. a. Plotting the data in Table 2-1 on log-probability paper. Duration Curves. Table 2-l shows the class intervals of daily flows input into the computer program STATS (58) for a duration analysis of Fishkill Creek at Beacon. Figure 2-2. New York.l . It has been found that making the logarithmic transform reduces the skewness of the curve.

42 86.0 0. or persistence.0 3b60.******.*. .95 6.0 306. the stream has a low ground-water storage and..**.0 2530.00 1500.30 15.5 54.EM 1110-2 1415 5 Mar 93 Table 2-1..*.00 99.**.********~****.00 98..*****. 3) water quality studies.*** l Lmm NUMBER PERCENT l LOWER Numm PERCENT l CLAPS ACCLM EQUAL OR l l l cuss NUHBER CLASS IN ACCUM l OR l l NUtBER .00 200.30 94.. NY . Es FLCW.00 15.00 50.*.5 1.CFS l EXCEED FLOW.00 3000. l l l l l l 5 6 7 4 2 3 5 4 228 37 66 95 25b 261 423 405 359 332 480 409 100.00 99.00 2000.60 14.**.0 l l l l l l l l l l l l l l 60.CFS l l *-------------------------------*-------------------------------* l l l l l l l * t l * l l * 0.**.00 8.00 4..**.00 600.. 2) determining minimum flow release.10 0.. The overall slope of the flow-duration curve is an indication of the flow variability in the stream.*****.********************.90 1000.00 .06 74.*.*. and 5) comparing yield potential of basins.47 41 9 3 3 0 0.. Daily Flow-Duration Data and Interpolated Values.00 6000. 1: 8 l l .******.**.00 30. Specific uses that have been made of duration curves are: 1) assessing the hydropower potential of run-of-river plants.00 5.61 99.0 3020.**.*********. NY .**.10 0.00 2.*.00 6. therefore.DAILY FLOWS *. u.------------------------------------------*------------------------------------------* .76 6.01 0.00 3..7 33.60 61.FISHKILL CR AT BEACON.. 1 1.*. b..67 44.00 4000.00 40...************.0 2.00 30.CFS CLASS NIJWER EXCEED 5606 4710 3935 2669 1667 1363 1043 570 339 66 EOUAL + l .*.0 656.00 300.*********.0 1500.00 95.0 171.36 97.8 99.09 66.95 l l l l l l l l l l l l l l l 5.0 1230.00 2.68 21. l l 11 12 13 14 15 15.94 l l l l l l l l l l l l l l l 16 17 16 19 20 21 22 23 24 25 26 27 20 29 30 100. If the lower end drops rapidly to the probability scale..***.03 0.50 99. 2-4 .00 20..CFS c&s NumER MCEED 6766 6761 6757 6749 6727 6690 6624 6529 6275 6014 7591 7166 6027 6495 6015 LMIT FLW.96 77.13 96.00 11.99 100.1 99..00 60..3 5.00 99.50 1.******.03 0.90 99.0 516.09 500. For some studies.00 .00 *********..*.00 118.00 60.62 30. a low or no sustained flow.20 0.7 l l l l l l l l l l l l l l 90.******* Output from HEC computer program STATS.0 1960..2 4.00 10.00 65.9 1..00 1.00 10.**.00 70.0 230.0 74..*****.****.00 0.00 600.00 80.4 22.0 903.0 420.DAILY FLOWS ***. may be more important (see Chapter 4).00 6970.***************.56 99.00 7000.***...40 91.00 400. It must be remembered that the chronology of the flows is lost in the assembly of data for duration curves.00 150.***. -DURATION DATA.90 99.*******. the low-flow sequence.59 3.00 20.9 1. . Duration curves are useful in assessing the general low flow characteristics of a stream.FISBKILL CR AT BEACON.**** PmcENT INTmwoLtrED l 1NTmLATED l l EQUAL OR MAGNITUDE l MAGNITUDE l l mKEED FLCW.00 668 763 1246 622 464 350 465 239 251 47 32 6 0 3 0 63.*********.00 .05 0. 4) sediment yield studies.00 40..62 53.00 SQ.00 50.***.**************************************** -INTERPOLATED DURATION CURVE.

d. 2-5 . Monthlv Curves. if there have been significant changes to the rating curve (because of major levee construction. Stage-Duration.EM 1110-2-1415 5 Mar 93 c. But. for instance) then the stage-duration curve should be derived from the flow-duration curve and the latest rating curve. Occasionally the distribution of the flows during particular seasons of the year is of interest. Stage-duration curves are often used to establish vertical navigation clearances for bridges. FISHKIU CREEK 10000 MAX IX 1000 E E 8 ii 5 c( d 10 100 5% 10% 30% 50% 70% 90% 95% 99% MIN 9 L ‘JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV OEC MONTH Figure 2-3. The log transformation is m recommended for stages. If there have been no changes in the discharge versus stage relationship (rating curve). Daily Flow-Duration Curves for Each Month. then the stages may be used instead of flows to compute a stage-duration curve. Figure 2-3 illustrates a way of presenting daily-flow-duration curves that were computed from the daily flows during each month.

derive a peak-flow curve by means of an empirical relation between mean daily flows and peak flows. or flood volume for a specified duration). conclusions drawn from information at the extremities can be misleading. and each event must result from a uniform set of hydrologic and operational factors. The transfer of data to either paper or a computer file always increases the probability that errors have been accidentally introduced. Errors have been noted in published reports of annual flood peaks. from the computed curve. Even though the observed data can be used to make inferences about future probabilities. 2-6 . Data selected for a frequency study must measure the same aspect of each event (such as peak flow. a. In such a case. 2-3. And. Occasionally. they should not be combined into a single series for frequency analysis. The data indicate there is a zero percent chance of exceeding 6970 cfs. but only the annual maximum events should be selected when the application of analytical procedures discussed in Chapter 3 is contemplated. maximum peak flows should be selected from the record. it is known that there is a finite probability of experiencing a larger flow. Figure 2-2 is based on 8766 daily values (Table 2-l). with newer records where the instantaneous peak was actually measured. Therefore. b. A duration curve is usually based on a fairly large sample size. there is some chance of experiencing a lower flow than the recorded 1. some other means is needed for computing the probabilities of infrequent future events. the entire record should be adjusted to a uniform condition (see Sections 2-3f and 3-9). For example. Future Probabilities. Uniformity of Data.le i n The primary question to be asked before selection of data for a frequency study is: how will the frequency estimates be used? If the frequency curve is to be used for estimating damage that is related to the peak flow in a stream. S 5s. errors have been found in the computer files of annual flood peaks. frequency studies are made separately for rain floods which occur principally during the months of November through March. For example. All reasonably independent values should be selected. if necessary. For instance. It is usually more reliable in such cases to segregate the data in accordance with type and to combine only the final curves. (1) Bm. care should be exercised when there has been significant change in upstream storage regulation during the period of record to avoid combining unlike events into a single series. (2) ms. it would be improper to combine items from old records that are reported as peak flows but are in fact only daily readings. Where two or more types of floods are distinct and do not occur predominately in mutual combinations. Selection of Data for Freauencv Analysis. Hydrologic factors and relationships during a general winter rain flood are usually quite different from those during a spring snowmelt flood or during a local summer cloudburst flood. For example. a reservoir’s behavior may be related to the 3-day or IO-day rain flood volume or to the seasonal snowmelt volume. where mean-daily flow records are more complete than the records of peak flows. the mean flow for several days’ duration may be appropriate.EM JIlO-2-1415 5 Mar 93 e. it may be more desirable to derive a frequency curve of mean-daily flows and then. it is necessary to select a related variable in lieu of the one desired. however. If the damage is best related to a longer duration of flow. mean-daily flow. Similarly. And similarly.I cfs. In the Sierra Nevada region of California and Nevada. Data should always be screened for errors. Section 2-4 describes the procedure for assigning probabilities to independent events.

the data should be adjusted as necessary to a single location. Estimating Missing Even& Occasionally a runoff record may be interrupted by a period of one or more years. and short-term climatic changes tend to be self-compensating. f. Some hydrologic records suggest regular cyclic variations in precipitation and runoff potential. In cases where no runoff records are available on the stream concerned. the frequency curve can be constructed for the stream-gage location and subsequently adjusted to the project location. their effect is ordinarily neglected. However. instead of attempting to estimate a complete series of individual events. Chapter 10 describes how to combine the separate frequency curves into one relation. For these reasons. the effect of natural cycles or trends during the analysis period is usually neglected in hydrologic frequency studies. Eff ect f R lationg. large areas that have been known to be fertile in the past are now arid deserts.those predominantly caused by snowmelt and those predominantly caused by rain.EM 1110-2-1415 5 Mar 93 and for snowmelt floods. design or operation of water resources control measures (structural and nonstructural). precipitation and. channel characteristics between the two locations must be taken into account. discussed in Chapter 9. failure to fill in the record for that flood would result in a biased data set and should be avoided. Although the existence of climatic changes is not questioned. it is often desirable to segregate hurricane floods from nonhurricane events. Climatic Cycles. summer thunderstorms should be excluded from frequency studies of winter rain flood or spring snowmelt floods and should be considered separately. which occur during the months of April through July. particularly the well established I l-year sunspot cycle. it can be used to extend the effective length of record at a location by adjusting frequency statistics (Section 3-7) or estimating missing events through correlation (Chapter 12). usually the location of the longer record. There is no doubt that long-duration cycles or irregular climatic changes are associated with general changes of land masses and seas and with local changes in lakes and swamps. Also. The anticipated effects of these measures in changing the rate and volume of flow is assessed by comparing the without project frequency curve with the corresponding with project frequency curve. and the recorded values must be adjusted to reflect uniform conditions in order that the frequency analysis will conform to the basic 2-7 . Flows for each of these two seasons are segregated strictly by cause . where appropriate. e. The differences in drainage area. d. because the long-term climatic changes have generally insignificant effect during the period concerned in water development projections. if the cause of the interruption is known to be independent of flow magnitude. In desert regions. it is usually best to estimate the frequency curve as a whole using regional generalizations. Where a longer or more complete record at a nearby station exists. and many attempts have been made to demonstrate that precipitation or streamflows evidence variations that are in phase with various cycles. (I) Hydrologic frequency estimates are often used for some purpose relating to planning. c. Location Differences. and large temperate regions have been covered with glaciers one or more times. and because of the difficulty in differentiating between stochastic (random) and systematic changes. Where data recorded at two different locations are to be combined for construction of a single frequency curve. When the stream-gage location is different from the project location. Along the Atlantic and Gulf Coasts. If the interruption is caused by destruction of the gaging station by a large flood. the record should be treated as a broken record as discussed in Section 3-2b. projects that have existed in the past have affected the rates and volumes of flows. Also.

Confidence limit curves should also be included for analytically-derived frequency curves to illustrate the relative value of the frequency relationships. along with the adopted frequency curves. adequate information should be included to permit an independent review of the data. (2) One approach to determining a frequency curve of regulated or modified runoff consists of routing all of the observed flood events under conditions of proposed or anticipated development. an annual-event curve and a partial-duration curve is shown on Figure 2-4. where there is considerable damage associated with the second largest and third largest floods that occurred in some of the years. A curve of annual maximum events is ordinarily used when the primary interest lies in the larger events or when the second largest event in any year is of little concern in the analysis. The selected series type should be established early in the study in coordination with the planner and/or economist. not cost effective. A summary of the basic data consisting of a chronological tabulation of values used and indicating sources of data and any adjustments should be included. Otherwise damage from the later flood would not be as large as computed. it is possible that analytical frequency analysis techniques can be applied. h. This type of curve is sometimes used in economic analysis. A map of the gage locations and tables of the adopted statistics should also be included. The text should indicate clearly the scope of the studies and include a brief description of the procedure used. Techniques of estimating project effects are outlined in Chapter 3-09d. Then a relationship is developed between the modified and the natural flows. routings of multiples of the largest floods of record or multiples of a large hypothetical flood can be used. the flows must be essentially unregulated by manmade storage or diversion structures. assumptions and analysis procedures. An ual Se ‘es Versus Partial Du at ion Seria There are two basic types of frequtncy c:rves uf:ed to estimate flood rdamage. care must be exercised to assure that the two are consistent. The time interval between flood events must be sufficient for recovery from the earlier flood. regardless of whether two or more occurred in the same year. A frequency curve of modified flows is derived from this relationship and the frequency curve of natural flows. Presentation of Data and Results of Freauencv AnalvsiS. ordinarily on probability paper. In order to meet the assumptions associated with analytical frequency analysis techniques. A graphic demonstration of the relation between a chronologic record. recorded runoff values should be adjusted to natural (unimpaired) conditions before an analytical frequency analysis is made. When both the frequency curve of annual floods and the partial-duration curve are used. because of the amount of work involved. 2-8 . wherever practicable. The partial-duration curve represents the frequency of all independent events of interest. The adjustment to natural conditions may be unnecessary and. In order to determine frequencies of runoff for extreme floods. When frequency curves are presented for technical review. Consequently. The frequency data should also be presented in graphical form. deriving an average or dependable relationship.EM 1110-2-1415 5 Mar 93 assumption of homogeneity. Caution must be exercised in selecting events because they must be both hydrologically and economically independent. including appropriate references. In cases where the impairment results from a multitude of relatively small improvements that have not changed appreciably during the period of record.

Illustration of Chronologic Sequence and Arrayed Flood Peaks. 1959-1868 9 8 7 6 5 4 3 2 1 0 1959 IQ60 IQ61 lQ62 1963 1964 1966 1966 1967 1966 Amy 10) d ” of Annual Flood Peakr. (Annual Seriee) lQ46-1968 I 9 8 7 6 5 4 3 2 1 0 1 5 9 10 15 20 24 Arrayed Sequence Number Amy of Flood Pe&e dove (Partial-Duration IIXJOda. 1945-1968 Serb) 10 a Q 7 6 . e 85 84 %3 a t 2 Ll 0 1 5 10 15 20 25 30 35 40 45 60 Amyed Sequence Number Figure 2-4. 2-9 .a 5 .EM 1110-2-1415 5 Mar 93 Chronological I” Sequence ofFloods.

that the derived curve can be easily visualized. The graphical method of frequency-curve determination can be used for any type of frequency study.(. Advantaaes and Limitations. That is. For these reasons. even though the frequency curves are obtained analytically. = 100 (1 . C. Every set of frequency data should be plotted graphically. Plot&-&&&. However. and that the observed data can be readily compared with the computed results.lOO(m -. General principles in the selection of frequency data are discussed in Section 2-3.. the plotting position. Also. graphical methods of frequency analysis are generally less consistent than analytical methods as different individuals would draw different curves. a. b. ated linearly between these two.4) where m is the order number of the event. Graohical Freauencv Analvsis. Comparison of the adopted curve with plotted points is not an index of reliability. but analytical methods ha\+: certain advantages when they are applicable. graphical procedures do not provide means for evaluating the reiiability of the estimates. stream or reservoir stages and regulated flows. . The principal advantages of graphical methods are that they are generally applicable.) of this value. This base value must be smaller than any flood flow that is of importance in the analysis. so that percent chance exceedance can be thought of as the number of events per hundred years. and should also be low enough so that the total number of floods in excess of the base equals or exceeds the number of years of record. and the corresponding plotting positions.3)/(N +. but it is often erroneously assumed to be. Graphical procedures should always be to visually check the analytical computations. N is taken as the number of years of record rather than the number of events. For arrays larger than 100. plotting positions that indicate more than one event per year can also be obtained using 2-10 (2-2b) . In ordinary hydrologic frequency work. Median plotting positions are tabulated in Table F. Selection and Arraneement of Peak Flow Data.1. the data arrayed in the order of magnitude. of the largest event is obtained by use of the following equation: P.EM 1110-2-1415 5 Mar 93 2-4. for example. thus implying a much greater reliability than is actually attained.5)1/N) (2-2a) The plotting position for the smallest event (P ) is the complement (1 -P. Table 2-2 is a tabulation of the annual peak flow data with dates of occurrence. graphical methods should be limited to those data types where analytical methods are known not to be generally applicable. The and all the other plotting positions are interpo P median plotting positions can be approximated by P. It is important to visually compare the observed data with the derived curve. particularly where there are more events than years (N). where frequency curves are too irregular to compute analytically. For partial-duration curves. P. Data used in the construction of frequency curves of peak flow consist of the maximum instantaneous flow for each year of record (for annual-event curve) or all of the independent events that exceed a selected base value (for partial-duration curve).

.. 1310.. -PLOTTING POSITIONS-FISEKILL CREEK AT BEACON.... 1760. . 2 3 3 1 2 2 3 3 5 27 15 16 1 8 1 12 25 13 20 16 10 21 11 6 26 13 28 26 9 15 30 19 1945 1945 1947 1946 1949 1950 1951 1952 1953 1954 1955 1955 1957 1957 1959 1960 1961 1962 1963 1964 1965 1966 1967 1966 2290.. f. 2.85 47... 2490. l .YZED..74 64.. Equation 2-2b.07 15. 980.15 60.. l l l l l l l l l l l l l l l l l 3020. .97 l l l l l 11. 2970...*. therefore.13 Output from HEC computer program HECWRC.. 2140.. N. Percent chance exceedance (or nonexceedance) is the recommended terminology. 1560... d. l l l 3 12 3 3 1 3 4 .... 1470..... The data plotted on Figure 2-5 shows a tendency to curve upward.44 72.. Unless computed by analytical frequency procedures. ...... 2220. This is simply an approximate method used in the absence of knowledge of the total number of events in the complete set of which the partial-duration data constitute a subset..... . 6260.. l . 2490.93 * l l l l l l l l l l 93..... 2500. A smooth curve should be drawn through the plotted points.... 4340.. The plotting grid used for stage frequencies is often the arithmetic normal grid. 1955 1956 1961 1968 1953 1952 1962 1949 1948 1956 1951 1945 1947 1960 1959 1963 1954 1967 1946 1964 1957 1950 1966 1965 8800. 3020.. This curve has been drawn through the plotted points. 4 l . 2500.46 31.. .. l 3 1 9 8 10 4 12 2 4 ...36 27.... ....03 97... 1960. ... Figure 2-5.....95 52. 1380.. **. WATER I-EEDIAU l l tBN DAY YEAR FLCU..ORDERED EVENTS. 3220. the frequency curve should be drawn as close to a straight line as possible on the chosen probability graph paper. ..05 56. s Exam le Plo tin The partial-duration curve corresponding to the partial-duration data in Table 2-3 is shown of Figure 2-6a. 1210. 2220. 1310....EM 1110-2-1415 5 Mar 93 Table 2-2.*... exceedance frequency...84 88.54 76.. . 3060.. .. ..87 6.-----------------------------------* l l . The plotting grids may contain a horizontal scale of exceedance probability.. . 1040... 2140.. 1580. Sequential and Arrayed with Plotting Positions.. 1470..*.56 35.. 8280. 3060....... . 3170. e... Annual Peaks. The annual-event curve was 2-11 ..34 66. .* . 3630.66 39....Y. ... a slightly curved line was drawn as a best fit line. . ExamDIe Plottim! of Annual-Event Freauencv Curve.. or percent chance exceedance.. 4340..... 3630...*..... 1380.FnNTS ANN.16 19.. 2290... Plottintz Grids. . Figure 2-5 shows the plotting of a frequency curve of the annual peak flows tabulated in Table 2-2.. .....CFS PLOT WS l t--------------------------. 1210... 1960... except that it was made to conform with the annual-event curve in the upper portion of the curve. ..75 * l l l l l l l 43. ..... + l l l l l l 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 16 19 20 9: 23 24 980. 1760. The plotting grid recommended for annual flood flow events is the logarithmic normal grid developed by Allen Hazen (ref 13) and designed such that a logarithmic normal frequency distribution will be represented by a straight line. 3170.. 1760. 1040.64 00. 1700. 2970.CFS l RANK YEAR FLoW.25 64.... 8800.. . 3220..26 23. ...

two-parameter gamma. When partial-duration data must include more events than there are years of record (see Subparagraph b) it will be necessary to use logarithmic paper for plotting purposes. the curve can be plotted on probability grid. I I 10 Exwrdmncr 1 .81 Chnom Figure 2-5.k lt I meon. log normal.ss 9s 0 I I E I I I Pehmnt I . extreme value (Gumbel) and log Gumbel.l . as illustrated on Figure 2-6b. a. Cimhkill 3-e. three-parameter gamma. Pearson type III.EM 1110-2-1415 5 Mar 93 developed in accordance with the procedures described in Chapter 3. 10? ss. log-Pearson type III. as on Figure 2-6a. Otherwise. Some theoretical distributions that have been used in hydrologic frequency analysis are the normal (Gaussian). Example of Graphical Frequency 2-12 Analysis . Chapter 3 describes the fitting of the log-Pearson type III to annual flood peaks and Appendix C describes fitting the extreme value (Gumbei) distribution. 2-5. exponential. estimating the parameters of the distribution from the data by some fitting technique. in order to plot exceedance frequencies greater than 100 percent. NY 46-‘1966 f 4 B 4 IL % d 4 : E a 103. and then evaluating the distribution function at various points of interest. The fitting of data by an analytical procedure consists of selecting a theoretical frequency distribution. General Procedures and Common Distributions. Analvtical Freauencv Analvsis.

. . ..cFs MEDIAN PLOT PCS l l NY -.. 04. .70 1967 1580... 16 .... .. . .. 52. .. . The use of an established procedure for fitting a selected distribution would result in consistent frequency estimates from the same data set by different persons..79 1963 51 ..93 1953 2080.92 33 1955 1660. .... 2-13 . .*.5 ABOVE 1500 CFS .72 31 1952 1730. .51 1956 1590. . .. . WATER NANKYEAN ....05 13 1951 2490.64 1948 19 2220.. 23 93...41 39 1600.. .33 1945 27 1780..46 1962 3060.**. . l .. ..CFS PLOT WS l g-----------------------------------* g-----------------------------------g . .. . .74 1951 20 2210.34 1953 2290. 64. The shape of this frequency curve would depend not only on the inflow but the capacities. .. . ..90 1951 42 1959 1560. Disadvantaees. ... WATER KEDIAN * .69 1948 50 1510.16 3630.. .. 19.. . . ..11 35 1650.. .. 35.62 32 1968 1720.. 80. 2 6.... 191. . . . 7 27. .... Therefore. l l .. . stage-frequency curves of annual maximum stages are shaped by the channel and valley characteristics.52 1956 29 1770. 43.. 195.... . 109. .' .. .. . .. .... Error distributions have been developed for some of the theoretical distributions that enable computing the degree of reliability of the frequency estimates (see Chapter 8). . . .04 1960 21 2140.. l l * . 1953 146. ..07 1955 8800.85 11 1949 2700.10 1955 1540. 150... . l l l l .. 60.36 6 1952 3170. .. 166. . For example. .62 30 1954 1760.... Arrayed with Plotting Positions.... . 121. 14 1952 . 39. . 23.21 36 1630.. 167. 199. .. operation criteria. b. NANKYEAN FLoW..39 1960 1520. FLcw. . .03 * . .... 134.. .. 1 2. 175.20 45 1951 1530. . 125. . . 1959 24 97.26 1953 : 3220. 129. .30 46 1968 47 1520. . . . .. Another example is the flow-frequency curve below a reservoir. 117. .25 15 1945 2290.95 12 1958 2500.. . .. .... 1959 1960. .. . 1968 15.. . . . . Partial Duration Peaks. .PEAK.44 1958 2290. etc. . .00 43 44 1550. .. . . FISEKILL CREEK AT BEACON. . 170. ..56 1949 8 3020. . .60 41 1560. . . 39 1956 158.EM 1110-2-1415 5 Mar 93 Table 2-3. ..23 1900.13 * 1920.. 31. ... .07 1961 4340. 207. 154.. .59 49 1952 1510.. 179. . . .02 34 1650.. AdvantageS.49 40 1956 1520... .66 9 1946 2970. .. 105. l . '66..31 37 1960 1610.. 22 88. ... 47. .. . The theoretical fitting of some data can result in very poor frequency estimates. 1952 26 1820. . . c.." .. . 1958 136. 1956 162...61 40 1580.. .. ..ORDERED EVENTS. . l l l l l l l l l l l l l l l l l l l l l l l . . . 72. graphical techniques must be used where analytical techniques provide poor frequency estimates.... . 1958 142. . ..54 1953 18 2280... Another advantage is that it is possible to regionalize the parameter estimates which allows making frequency estimates at ungaged locations (see Chapter 9). 56. backwater conditions. ..75 1948 10 2750. 113. 17 .. 203. .. ..97 1956 8280. ..43 1963 20 1760. 3 11.. 1958 25 101. .15 2460. 183. . . 76. etc.. . Determining the frequency distribution of data by the use of analytical techniques has several advantages. . . ORDERED EnNTS..

Probability Paper. 2-14 .99 99. Log-Log Paper I . Partial Duration Frequency Curve..9 9s mr SO of 2Wrlt.1 .~ 60 pn lalneme 18 Y9m-9 1 . Partial Duration Frequency Curve. i M L :: h 1r* 99.EM 1110-2-1415 5 Mar 93 mr of 2HOtm pwr lwnerrd YW'rrt-9 Figure 2-6a.l Figure 2-6b.

Fitting the Distribution. variance. (estimated by the sample variance. Various aspects of the procedures are described in an attempt to clarify the computational steps. 3-2. (1) The log-Pearson type III distribution is fitted to a data set by calculating the sample mean. Since the distribution is a logarithmic distribution. or third moment. The parameters are: the mean. it must be recognized that occasionally. or second moment.” Therefore. the skew. The Pearson type III distribution is particularly useful for hydrologic investigations because the third parameter. G). rather than from the observations themselves. permits the fitting of non-normal samples to the distribution. X). As stated in Bulletin 17B.. all parameters are estimated from logarithms of the observations. or first moment. Bulletin 17B (46). When it is necessary to use a procedure that departs from Bulletin 17B. a. (estimated by the sample skew.. S*). When the skew is zero the log-Pearson type III distribution becomes a two-parameter distribution that is identical to the logarithmic normal (often called log-normal) distribution. ANALYSIS The procedures that federal agencies are to follow when computing a frequency curve of annual flood peaks have been published in Guidelines for Determining Flood Flow Frequency. This distribution requires three parameters for complete mathematical specification. the variance. This report contains most aspects of Bulletin 17B. “Flood events .EM 1110-2-1415 5 Mar 93 CHAPTER 3 FLOOD FREQUENCY 3-I. the procedure should be fundamentally sound and the steps of the procedure documented in the report along with the frequency curves. the recommended techniques may not provide a reasonable fit to the data. General. The step by step procedures to compute a flood peak frequency curve are contained in Appendix 12 of Bulletin 17B and are not repeated herein. b. Introduction. Lon-Pearson TVDe 111Distribution. (estimated by the sample mean. The intent herein is to provide guidance for use with Bulletin 17B. and skew from the following equations: Z7- P N (3-l) 2 1x2 S z-c N-l 1 (xx)* (3-2a) N-l 3-l . The analytical frequency procedure recommended for annual maximum streamflows is the logarithmic Pearson type III distribution. the skew. do not fit any one specific known statistical distribution. but in an abbreviated form.

while for larger skew values all of the values in the table would ordinarily be needed. S) represents the slope of the curve.EM 1110-2-1415 5 Mar 93 0 1 X2 .l)(N-2)S3 ST = mean logarithm X = logarithm of the magnitude of the annual event N = number of events in the data set S2 X unbiased estimate of the variance of logarithms = X-z.3N(l X)(1 X2) + 2(cX)3 (3-3b) N(N. For a skew value of zero.I)(N-2)S3 5 N(C (X-y)3) (N. The number of values needed to define the curve depends on the degree of curvature (i. the mean represents the general magnitude or average ordinate of the curve.1. the deviation of the logarithm of a’ single event from the mean logarithm unbiased estimate of the skew coefficient of logarithms digits G= The precision of the computed values is more sensitive to the number of significant when Equations 3-2b and 3-3b are used. Computation of the unadjusted frequency curve is accomplished by computing values for the logarithms of the streamflow corresponding to selected values of percent chance exceedance.(1 X)*/N N-l (3-2b) G= WC x3) (N. the skew). (3) The logarithms of the event magnitudes corresponding to each of the selected percent chance exceedance values are computed by the following equation: IogQ = ?t+KS (3-4) where ‘jTand S are defined as in Equations 3-1 and 3-2 and where 3-2 .e..l)(N-2)S3 (3-3a) m in which: N*(c X3) . and the skew represents the degree of curvature. only two points would be needed. (2) In terms of the frequency curve itself. the square root of the variance (the standard deviation. A reasonable set of values and the results are shown in Table 3.

...* PERCEETl . l 1100.. 48.3684 + 2. FREQUENCY CURVE STATISTICS mANLoGARIlm STANDARD DEVIATION comuIEDsKEw l STATISTICS BASED ON 0 l t--------------------------------. Table 3-l and Figure 3-l shows the derived frequency curve along with the expected probability adjusted curves and the 5 and 95 percent confidence limit curves.**. ExamDie Computation.. l 3650. 14500.. ...CONFIDENCE LIMITS. . .. l 99...**... nPEcTED l CEANCE l l CCMWTED PROBABILITY l EKEEDANCE * 0. 7100.O...3604 0.* . .------------------------. 804. 1200. S and G are taken from Table 3-1. . c. . where X.....0 l 1110..... l 6850..11500 cfs (2) It has been shown (36) that a frequency curve computed in this manner is biased in relation to average future expectation because of uncertainty as to the true mean and standard deviation. Equation 3-4 is solved by using the computed values of X and S and obtaining from Appendix V-3 the value of K corresponding to the adopted skew.... l 10.5 l 9740. l 6640. ... . This adjustment is discussed in more detail in Section 3-4. . The effect of this bias for the normal distribution can be eliminated by an adjustment termed the expected Dadiustment that accounts for the actual sample size.*.. l 90.05 LIMIT 0. * 95.. 1040. l 80.. .. 7820.95 LIMIT * . 1010.0 * 1790. * 0.0 l 3950.0 * 746. . 11500. NEW YORK ..6000 0... ....**.0 l 10800... .. 14100.EM 1110-2-1415 5 Mar 93 log Q = logarithm of the flow (or other variable) corresponding of percent chance exceedance K to a specified value = Pearson type III deviate that is a function of the percent chance exceedance and the skew coefficient..****.***.. l 4.. * 8080. 28300. is: log Q = 3. ..0 l 4710. 2990...... 1170.. .2456 0. * * l l l GENERALIZEDSKEW ADOPTED SKEW 3.------------------------. ....2456) = 4.. l 20. l 2650. . 5210. 2 TI - isti -FRXQUENCY CURVE. l 1..... (I) As shown iL the following example. . l ...7300 0. An example computation for P=l. ..*.7000 l l l l l HISTORIC EVENTS Ena OUTLIEXS mouTLIERS ZERO OR MISSING l 0 0 0 * l l SYSTEHATIC EVENTS 24 l . 3-3 . l 0. 791.**. .. 19200. ..0 l 14800. 10500..2 l 39100.. . andthe selected percent chance exceedance (P).. 9110. 3740.. .*.------------... 19000..01-3735 FISEKILL CREEK AT BEACON. 4960.. ...0 l 1490.. * 1320.. .. * 1760..***. l . l 26900.. l 12300...... 5380.0619 Q . . 641.*.. l 50.0 l 20100. 1440. .. 2190..CPB.-----------------------------.. l 2. 1420. G..*. 2190...... .0 l 568.8236 (. .FLW.

The gage may have been temporarily discontinued for budgetary or other reasons. Different analysis procedures are recommended for missing high events and for missing low events. Missing high events may result from the gage being out of operation or the stage exceeding the rating table. Annual Frequency Curve. 7 C 6 Y re3. e. bottom of a crest stage gage. An incomplete record can result when some of the peak flow events were either too high or too low. 3-4 . Broken Rem . d. the missing events were caused by a random occurrence.1. In these cases. HY ! 0 Obeened Computed - Annual Peakr Curve Adjurtment - Frequency With Expected Confidence Probability Curwe Limit :0 s ii P z : 0 4 led-- / =I .01 Figure 3.EM 1110-2-1415 5 Mar 93 Fishhill Crrrk rt Or&on.- 98 PIrrrnt Chrncr 60 Excrrdmcr l8 _- I‘ . In these cases. then the observed events should be analyzed as an incomplete record. InCOmDlete Record. Missing low floods usually result when the flood height is below the minimum reporting level or the . If a portion of the record is missing because the gage was destroyed by a flood or the flood was too low to record. The different segments of the record are added together and analyzed as one record. the record should be analyzed using the conditional probability adjustment described in Appendix 5 of Bulletin 17B and Section 3-6 of this report.I . In other words. A broken record results when one or more years of annual peaks are missing for any reason not related to the flood magnitude. unless the different parts of the record are considered non-homogeneous. every effort should be made to obtain an estimate of the missing events.

Zero-flood years. When the skew coefficient is between -0. = X + K. Use plus value for high-outlier threshold and minus value for low-outlier threshold N= Sample size (may be historic period (H) if historically adjusted statistics are used) (3) Hiah Outliers.4 and +0. Flood peaks that are above the upper threshold are treated as high outliers. (1) Guidance. When the skew coefficient is less than -0.4. The following equation is used to screen for outliers: (3-5) J7. either historical information must be available or the probable occurrence of the event(s) estimated based on flood information at nearby sites. Outliers. for any flood peak(s) to be weighted as high outlier(s). then the outlier(s) should be retained as part of the systematic record.4.S where: X0 = outlier threshold in log units ST = mean logarithm (may have been adjusted for high or low outliers. the high-outlier test is applied first and the adjustment for any high outliers and/or historic information is made before testing for low outliers. the low-outlier test is applied first and the adjustment for any low outlier(s) is made before testing for high outliers and adjusting for any historic information.4. Some of the gaging stations in arid regions record no flow for the entire year. If it is not possible to obtain any information that weights the high outlier(s) over a longer period than that of the systematic record. In this case the record should be analyzed usinn the conditional probability adjustment described in Appendix 5 of Bulletin 17-Band Section 3-6 of this report. (2) bation. A zero flood peak precludes the normal statistical analysis because the logarithm of zero is minus infinity. When the computed (station) skew coefficient is greater than +0. = K value from Appendix 4 of Bulletin 17B or Appendix F.EM 1110-2-1415 5 Mar 93 f. 3-5 . g. The one or more values that are determined to be high outliers are weighted by the historical adjustment equations.” The sequence of steps for testing for high and low outliers is dependent upon the skew coefficient and the treatment of high outliers differs from that of low outliers. and/or historical information depending on skew coefficient) S= standard deviation (may be adjusted value) K. Therefore. The Bulletin 17B (46) defines outliers as “data points which depart significantly from the trend of the remaining data. both the highand low-outlier tests are made to the systematic record (minus any zero flood events) before any adjustments are made. Table 11 of this report.

(1) Definitions. Historical information is knowledge that some flood peak. 3-3. a. that the skew coefficient computed from a small sample is highly unreliable. It can be demonstrated. it may be desirable to test the sensitivity of the results by considering the value(s) as low outlier(s). Appendix D in this manual is a reprint of Appendix 6 from Bulletin 17B and contains the equations for adjusting for historic events and/or historical information. + MSE. Bulletin 17B recommends the following weighting equation: MSEE(G) + MSE. the skew coefficient computed from a small sample may depart significantly from the true skew coefficient of the population from which the sample was drawn. but above the threshold value.) b. Low outliers are deleted from the record and the frequency curve computed by the conditional probability adjustment (Section 3-6). That is. h. the skew coefficient must be compared with other representative data. Also a basic assumption in the adjusting equations is that no peaks higher than the lowest historic event or high outlier occurred during the unobserved part of the historical period. because every peak in the systematic record that is equal to or larger than the lowest historic peak must be treated as a high outlier. was the largest event over a period longer than that of the systematic record. where: (3-6) Gbl = c MSEi MSE. Weiahted Skew Coefficient. General. Flood peaks that are below the low threshold value are treated as low outliers. A more reliable estimate of the skew coefficient of annual flood peaks can be obtained by studying the skew characteristics of all available streamflow records in a fairly large region and weighting the computed skew coefficient with a generalized skew coefficient. Weinhtine Eauation. (2) Eauations. The adjustment equations are applied to historic events and high outliers at the same time. It is important that the lowest historic peak be a fairly large peak.(@ G. If there are one or more values very near. Historic events are large flood peaks that occurred outside of the systematic record. either through the theory of sampling distributions or by sampling experiments. = MSE. (Chapter 9 provides guidelines for determining generalized skew coefficients. Consequently.EM 1110-2-1415 5 Mar 93 (4) La. weighted skew coefficient G = computed (station) skew = generalized skew = mean-square = mean-square error of generalized skew error of computed (station) skew 3-6 . either systematic or historic. Historic Events and Historical Information. It is historical information that allows a high outlier to be weighted over a longer period than that of the systematic record.

30 IGIif IGl > 0. These two error distributions are combined in the formation of the non-central t distribution. a. Mean Sauare Error.90 B = = where: IQ 0. 3-4. For the normal distribution. the MSE would be the average of the squared differences between the computed (station) skew coefficients and the isoline values. And. (Chapter 8 discusses the reliability and the distribution of the computed statistics. if an arithmetic mean of the stations in a region were adopted. Exoected Probability.) The fact of not knowing the location of the true frequency curve is termed uncertainty. encompass the true frequency curve. For a prediction equation. Figure 3-2 shows 3-7 . the square of the standard deviation (variance) would approximate the MSE.26 IGI if IGlz 0.50 = absolute value of the computed skew N = record length in years Appendix F.94 . (1) The mean-square error of the computed skew coefficient for log-Pearson type III random variables has been obtained by sampling experiments. = = A 10CA-BCLog~~W~~)l~ (3-7a) (3-7b) 0. The non-central t-distribution can be used to construct curves that.10 provides a table of mean-square coefficients based on Equation 3-7a. with a specified confidence (probability).0. For an isoline map. as an estimate of the distribution parameters.90 10A+B/NB = -0. the sampling errors for the mean are defined by the t distribution and the sampling errors for the variance are defined by the chi-squared distribution. Equation 6 in Bulletin 17B provides an approximate value for the mean-square error of the computed (station) skew coefficient: MSE. error for several record lengths and skew (2) The mean-square error (MSE) for the generalized skew will be dependent on the accuracy of the method used to develop generalized skew relations. the square of the standard error of estimate would approximate the MSE.33 + 0. The computation of a frequency curve by the use of the sample statistics.55 SO if IGl > 1.52 + 0.08 IGl if IGlf = -0. provides an estimate of the true frequency curve.EM 1110-2-1415 5 Mar 93 c.

5 percent chance that the “true” curve would indicate 0. If one wished to design a flood protection work that would be exceeded.Ol Figure 3-2. would actually result in an average of 2. . Figure 3-3.25 4 . .. the usual design would be based on the normal standard deviate of 2.ix F-8). there is a 99.EM 1110-2-1415 5 Mar 93 the confidence li_mit curves around a frequency curve that has statistics: N= IO.975 .326. A graph of the number of possible “true” exceedances versus the probability that the true curve exceeds this value. on the average. c.69 exceedances (see Apperi .9 99 P&NT CHIN&! EXCEED& 1 . the following assumed b.O.990 . SAMPLE SIZE-10 ERROX LIMIT EXCEEDAYCL lXDO. NORMAL 5 OISTRIBUTION.l . only one time every 100 years (one percent chance exceedance).5 percent chance that this design level may come from a “true” curve that would average 22 exceedances per 100 years. X=0. provides a curve with an area equal to the average (expected) number of exceedances. Note the large number of exceedances possible on the left side of the curve.90 .995 2 1 90. Notice that there is a 0. The design of many projects with a target of 1 exceedance per 100 years at each project and assuming N-IO for each project.75 d L z 9 B . = 5 a’ 2 - 3 . Confidence Limit Curves based on the Non-central 3-8 t Distribution.50 . This relationship is highly skewed towards the large exceedances because the bound on the right side is zero exceedance.950 .004 exceedances.99 99. On the other side of the curve. instead of the expected one exceedance. S= 1.

EM 1110-2-1415 5 Mar 93

TARGET AN0 USED

Of

1 ON

LXCEEDAYCE SAMPLE SIZE

PER OF

100 10

EVENTS EVEYlS

Figure 3-3. Cumulative Probability Distribution

of Exceedances per 100 Years.

d. There are two methods that can be used to correct (expected probability adjustment) for this bias. The first method, as described above, entails plotting the curve at the “expected” number of exceedances rather that at the target value, drawing the new curve and then reading the adjusted design level. Appendix F-8 provides the percentages for the expected probability adjustment. e. The second method is more direct because an adjusted deviate (K value) is used in Equation 3-4 that makes the expected probability adjustment for a given percent chance exceedance. Appendix F-7 contains the deviates for the expected probability adjustment. These values may be derived from the t-distribution by the following equation: K P,N where: P = exceedance probability (percent chance exceedance divided by 100) N= Kt sample size expected probability adjusted deviate = Student’s t-statistic from one-tailed 3-9 distribution

[tN+l t P,N-1 PI”

=

(3-8)

EM 1110-2-1415 5 Mar 93 f. For a sample size of 10 and a 1% percent chance exceedance, the expected probability adjusted deviate is 2.959 as compared to the value of 2.326 used to derive the computed frequency curve. g. As mentioned in the first paragraph, the non-central t distribution, and consequently the expected probability adjustment, is based on the normal distribution. The expected probability adjustment values in Appendices F-7 and F-8 are considered applicable to Pearson type III distributions with small skew coefficients. The phrase “small skew coefficients” is usually interpreted as being between -0.5 to +0.5. Note also that the uncertainty in the skew coefficient is not considered. In other words, the skew coefficient is treated as if it were the population skew coefficient. h. The expected probability adjustment can be applied to frequency curves derived by other than analytical procedures if the equivalent worth (in years) of the procedure can be computed or estimated. 3-5. &j&. a. Definition. The term risk is usually defined as the possibility of suffering loss or injury. In a hydrologic context, risk is defined as “the probability that one or more events will exceed a given flood magnitude within a specified period of years” (46). Note that this narrower definition includes a time specification and assumes that the annual exceedance frequency is exactly known. Uncertainty is m taken into account in this definition of risk. Risk then enables a probabilistic statement to be made about the chances of a particular location being flooded when it is occupied for a specified number of consecutive years. The percent chance of the location being flooded in any given year is assumed to be known. b. Binomial Distributioq. for the binomial distribution:

R, s N!

The computation

of risk is accomplished

by the equation

I!(N-I)!

P’(l-P)N-*

U-9)

where:

RI =

N=

risk (probability)

of experiencing

exactly I flood events

number of years (trials)

I = number of flood events (successes) P = exceedance probability, percent chance exceedance divided by 100, of the annual event (probability of success) (The terms in parentheses are those usually used in statistical texts) When I equals zero (no floods), Equation 3-9 reduces to:

3-10

**EM 1110-2-1415 5 Mar 93
**

RO

= (1-P)”

(3-10a)

and the probability of experiencing one or more floods is easily computed by taking the complement of the probability of no floods: R (1 = 1-(1-P)” (3-lob)

or more)

c. Aoolication. (1) Risk is an important concept to convey to those who are or will be protected by flood control works. The knowledge of risk alerts those occupying the flood plain to the fact that even with the protection works, there could be a significant probability of being flooded during their lifetime. As an example, if one were to build a new house with the ground floor at the 1% chance flood level, there is a fair (about one in four) chance that the house will be flooded before the payments are completed, over the 30-year mortgage life. Using Equation 3- lob: R (1 = 1-(1-.01)3a = l-.9930

= l-.74

or more)

= .26 or 26% chance

.

(2) Appendix F-12 provides a table for risk as a function of percent chance exceedance, period length and number of exceedances. This table could also be used to check the validity of a derived frequency curve. As an example, if a frequency curve is determined such that 3 observed events have exceeded the derived 1% chance exceedance level during the 50 years of record, then there would be reason to question the derived frequency curve. From Appendix F-12, the probability of this occurring is 0.0122 or about 1%. It is possible for the situation to occur, but the probability of occurring is very low. This computation just raises questions about the validity of the derived curve and indicates that other validation checks may be warranted before adopting the derived curve. 3-6. Conditional Probabilitv Adiustment. The conditional probability adjustment is made when flood peaks have either been deleted or are not available below a specified truncation level. This adjustment will be applied when there are zero flood years, an incomplete record or low outliers. As stated in Appendix 5 of Bulletin 17B, this procedure is not appropriate when 25 percent or more of the events are truncated. The computation steps in the conditional probability adjustment are as follows: 1. Compute the estimated probability <F>that an annual peak will exceed the truncation level:

F = N/n

3-11

(3-l la)

the exceedance frequency for that flow can be estimated.KS. The computed frequency curve is actually a conditional frequency curve.. 3.dQso) Gs = -2.. Q and Qso the discharges determined in Step 3..) log (Q. S. 2. 10 and 50 percent chance exceedances.. The conditional exceedance frequencies are converted to annual frequencies by the probability computed in Step 1:. If the statistics reflect the adjustments for historic information. S. 6.12 (3./Q. (3-15) where G.) for 1. respectively.50 + 3.EM 1110-2 1415 5 Mar 93 where N is the number of peaks above the truncation level and n is the total number of years of record.13) ss = log (Q.../Qso) Kl-K50 (3-14) Xs = log (Qso) . Estimate log-Pearson type III statistics that will fit the upper portion of the adjusted curve with the following equations: log (Q. 3-12 . standard deviation and mean. P = FPd where P is the annual percent chance exceedance and P.WL F= H where H is the length of historic period. Interpolate either graphically or mathematically (Qt. Given that the flow exceeds the truncation level. and K and K are the Pearson Type ’f II deviates for percent change exceedances of 1 and 30 ansg skew coefficient G. W is the systematic record weight and L is the number of peaks truncated. (3-12) percent (3-l lb) to obtain the discharge values 4. and z are the synthetic skew coefficient. then the appropriate equation is H . is the conditional chance exceedance.

it can be a significant problem if it becomes necessary to estimate more than a few events.EM 1110-2-1415 5 Mar 93 5. runoff or precipitation there are locations over a long period. While this may not be a serious problem if only one or two events must be estimated to “fill in” or complete an otherwise unbroken record of several years.(~)2/Nl = the determination coefficient station station Y = the value at the short-record X = the concurrent value at the long-record N = the number of years of concurrent record 3-13 . The longer record station selected for the adjustment procedure should be in a hydrologically similar area and. missing events should not be freely estimated by regression analysis. Purnose. (2) using the computed degree of correlation and the statistics of the longer record station to compute an adjusted set of statistics for the shorter-record station. Develop the computed frequency curve with the synthetic statistics and compare it with the plotted observed flood peaks. Combine the synthetic skew coefficient with the generalized skew by use of Equation 3-6 to obtain the weighted skew. correlation coefficient [BY The degree of correlation is reflected in the R2 as computed through use of the following equation: . if possible. (2) It is possible. in frequency studies. and (3) computing an equivalent “length of record” that approximately reflects the “worth” of the adjusted statistics of the short-record station. by regression or other techniques. to estimate from concurrent records at nearby locations the magnitude of individual missing events at a station. Cm. (I) In most cases of frequency studies of in the region where records have been obtained of record at such a nearby station is useful for station provided there is reasonable correlation locations. (3) The procedure for adjusting the statistics at a short-record station involves three steps: (I) computing the degree of correlation between the two stations. have a drainage area size similar to that of the short-record station.(~~VN12 R2 3: (3-16) [B2 . 3-7. The additional period extending the record at a short record between recorded values at the two Comoarison. the use of regression analysis produces estimates with a smaller variance than that exhibited by recorded data. Consequently. 6.(fW2/Nl where: R2 rp* . Two-Station a. b. However.

the standard deviation: The following equation can be used to adjust s. to use the logarithms of the c.z= Sy: +cS:$1 R*(S. .I R (s.) where: (3-17) T = the adjusted mean at the short-record P. e. Adiustment of Mean.17 should not be applied.~ (approximate) 3-14 . for annual flood peaks.. Equation 3.. If R2 is less than the criterion./S.2) (3-18) where N. d. See Appendix 7 of Bulletin l7B for procedures to compare the variance of the adjusted mean against the variance of the entire short-record period. = the mean for the concurrent R= S S Yl record at the longer-record the correlation coefficient = the standard deviation for the concurrent = the standard deviation for the concurrent station record at the short-record record at the longer-record station Xl AI1 of the above parameters may be derived from the logarithms of the data where appropriate.. The criterion for determining if the variance of the adjusted mean will likely be less than the variance of the concurrent record is: R* > l/(N. + (X3 . = the mean for the concurrent station station station station record at the short-record J7. it is appropriate values in the equations in this section.EM 1110-2 1415 5 Mar 93 For most studies involving streamflow values./S.%.g. In this case just use the computed mean at the short-record station or check another nearby long-record station. The following equation is used to adjust the mean of a short-record station on the basis of a nearby longer-record station: T = T.. equals the number of years of concurrent record. = the mean for the complete record at the longer-record K. Adiustment of Standard Deviation.

= the number of years of record at the longer-record station R= the adjusted correlation coefficient Figure 3-4 shows the data and computations for a two-station comparison for a short record station with 21 events and a long record station with 60 systematic events. f. where: NY= the equivalent length of record of the mean at the short-record station = the number of years of concurrent record at the two stations Ny1 N.19 in Appendix 7 of Bulletin 17B. but in most cases the difference in the results does not justify the additional computations. When adjusting the statistics of annual flood peaks either a weighted or a generalized skew coefficient may be used depending on the record length.. e. e.(1 ..] [R2 . There is no equation to adjust the skew coefficient that is comparable to the above equations. 1 . 3-15 . Adiustment of Skew coefficient. It can be seen that the adjustment of the frequency statistics provides an increased reliability in the mean equivalent to having an additional 17 years of record at the short-record station. mRecord The final step in adjusting the statistics is the computation of the “equivalent record length” which is defined as the period of time which would be required to establish unadjusted statistics that are as reliable (in a statistical sense) as the adjusted values.W. - (3-20) .3)l N. the equivalent length of record is an indirect indication of the reliability of the adjusted values of Y and S. for annual flood peaks. The equivalent record length for the adjusted mean is computed from the following equation: NY = NY.EM 1110-2-1415 5 Mar 93 where: SY= S Yl the adjusted standard deviation at the short-record station record at the short-record = the standard deviation for the period of concurrent station SX = the standard deviation for the complete record at the base station = the standard deviation for the period of concurrent = the determination coefficient record at the base station All of the above parameters may be derived from the logarithms of the data where appropriate.. Thus.g.. This equation provides approximate results compared to Equation 3.)/N.R2VW.

0 95.. Figure 3-4.0811 MSES = 0.0 SO..?90495 Correletion Coefficient: From equrtions I1 = 3. 3-16 . peek Largest since 1~).1 FREOUEYCY CURVE.811351 (See Figure P-00 6870 19. .7340 YOC ^. 21.0811 -0.o.~0... - cm: Slope: b = 0.-.4445 $1 2 = t(0.142 i.23M = = 38 1 . end 22: * 0. .-e-esystematic Events -4 I rota1 Concurrent -_-__------_---_-____ 21 40 71 0 3. TALLULAH RIVER NEAR CLAYTON.8450 0. CDMPUTED -95 LIMIT 9580 8410 7540 4480 5550 4480 3780 2420 1450 1060 847 524 7850 Nistorjc inforution. = 0. EXPECTED PROBAIILITY 127Do 10900 9590 8350 4740 Efi! 2810 1760 1370 1110 745 142OD 11900 10300 8800 4990 44aD 5710 2610 1740 1340 1070 406 PERCENT CNANCE EXCEEDANCE :: i-8 s:o 20.. = -0. l (0..CFS.0 .?=+ .2322 Skew Computed 0.0 % 7440 5140 2800 3100 2470 2010 976 2160 8500 4660 2410 4530 3580 4090 4240 2880 1400 19bD 3260 2000 1010 c.. .. = 3.2409 -0..1588 concurrent 39 0 3.302 l Ii n (0. .11114)2~0.9310) Y. I ______ 21 0 3.0 10.302)(-0.4794 R =0.2678)21(0..24713 0.M6b l 3.?90495 (3..1844 (0. = NSEi = 0.2b?8~2 9 f 1 "I( (0.0111) .5075 I ..COYFIDENCE .2140 0.055 =* -0.EM 1110-2-1415 5 Mar 93 Iear -r--- flow - I 1915 1917 1918 1919 1920 1921 1922 1923 192L 1925 1924 1927 1928 1929 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 19SL 19% 1956 1957 1958 1959 1960 1961 1962 1963 1964 1945 1966 1967 196a I969 1970 I971 I972 1973 1974 I975 1974 I977 I978 I979 I980 I981 I982 1983 I984 I965 l Chettooge River 1ellul.0:63 -0.0142)(0. 7530 4870 3840 2930 4450 6440 12400 13900 4740 5220 13400 4020 4230 5820 5820 5620 5420 2% 7310 9660 5420 9880 27200 13400 15400 5620 14700 3480 3290 7440 19400 4400 6340 18500 13000 14800 lD900 4120 5000 7910 4810 4740 .::a3 ..L794 0.os LIMIT 18100 15400 13400 11300 8810 7040 5370 3260 2060 1440 1340 957 GA LIMITS.' .h River 12400 14000 5900 14000 8200 4100 4200 5300 9200 3900 4200 3400 20100 11400 29000.0 99.2b09~2 = 0. Two-Station Comparison Computations.FLDM. .0) ..“.0 80.2b09/0.9310 0.1000 Wistofic Period Log *em Standard Deviation Cenerelised Adopted 13.2322)2 .0 90.

Although N is actually the equivalent years of record for the mean.) Figure 3-5 shows the resulting unadjusted and adjusted frequency curves based on the computed and adjusted statistics in Figure 3-4.. the value is used as an estimateYequivalent record length in the computation of confidence limits and the expected probability adjustment.s 9s I I I Hill I I I III Hlll 9-0 Pmrcmnt Chmer se Erardmcr 1-e siss i . station.i . station. station using Equations 3. .ir Figure 3-5. for the short-record ---- Expected Probability Probability Curve Cunc from Data from Two-Station Comrwiaon Exwctcd 1e4 103. for the entire record at the longer-record (4) Compute %. Summarv of Stem. station which is (6) Compute y and S. it?=- liiiiilii 9s’. (5) Compute the correlation coefficient using Equation 3. for the entire record at the short-record (3) Compute % and S. and S. for the portion of the longer-record concurrent with the shojt-record station. The procedure for computing and adjusting frequency statistics using a longer-record station can be summarized as follows: (1) Arrange the streamflow data by pairs in order of chronological sequence. and S. Observed and Two-Station 3-17 Comparison Frequency Curves. (2) Compute y. g.EM I I lO-2-1415 5 Mar 93 (Figure 9-1 shows the computations for b and R and Figure 9-2 shows the Tallulah River annual peaks plotted against the Chattooga River peaks.16.17 and 3.18.

8 1080.6 396.7 706.0 1547.7 737.0 307.4 654.7 586.0 1616.4 524.3 876.7 811.5 296.6 872. Flood volume-duration data normally obtained from the USGS WATSTORE files consists of data for 1.7 324.1 820.7 900.2 451.0 463. Table 3-2.7 654.7 3354. and 183 days.9 1951 2160.1 1023.VOLUME-DURATION DATA .4 1714.7 88B.8 450.7 235.0 1247.8 460.3 1248.S 1950 1oso.7 1309.5 346.0 1106.7 352.0 2233.0 1949 2900.1 455.0 305.3 429.5 906.6 1960 2080. 30.6 522.0 1947 1800.5 782.8 1958 2130.0 848.0 430.7 1670. NY .9 1957 1230.0 405.9 1955 6970.DAILY FLOWS ----------------------------------------------------------------------------------------------BIGHi? 7 bEAN VALUE FOR DURATION.5 1954 1096.5 1948 2660.2 777.2 1962 2570.7 437.5 3500. Whenever it is necessary to consider flows separately for a portion of the water year such as the rain season or snowmelt season. Flood Volumes.1 2546.4 325.5 1032.6 466.1 517.3 19s9-7 767.1 1952 2670.4 372.2 636.6 1159.4 1145.7 600.1 1965 826.3 1961 3440.8 1964 1300.2 994.7 1398. These same values are the default values in the HEC computer program STATS (Table 3-2).o 461. (9) Compute the expected probability adjustment 3-8.7 641.5 657.0 837.2 460. 90.3 175.3 1037.0 1770.3 605.2 759.7 567.0 652.3 445.3 1317.1 712.6 701.1 1959 986. Flood volume-duration curves are used primarily for reservoir design and operation studies.9 692.1 470.7 422.6 569.2 1956 6760.0 5456.8 752.0 1046.S 90 30 60 15 TEAR ----------------------------------------------------------------------------------------------- 1 3 120 183 559.0 900.0 S46.1 611.3 645.8 486.2 408.5 687.0 2430.7 1360.1 1968 2810.1 1354.6 303. and should generally be developed in the design of reservoirs having flood control as a major function.FISIiKILL CR AT BEACON.1 759. FL0w.3 455.9 396. the same durations (up to the 30-day or 90-day values) are selected from flows during that season only.7 756.3 541. Hinh Flow Volume-Duration Data .9 680.3 605.7 611.0 1328.0 2346.7 1520. 0 909.5 6Q3.0 1616.6 1557.1 370.4 521.8 1966 774. 15.5 1967 1416.9 568. 60.2 746.6 945.0 1934.4 1715.CF.2 2155.4 512.1 781.0 1641.3 373.0 436.5 609.1 560.2 514.7 lS90.4 491.4 335.1 2322.5 656.0 209.6 925.1 162.0 4536.7 1520.9 363.4 857.0 1160.7 639.7 397.3 1106.3 923.0 620.7 635.3 1945 2080.3 275.6 631.7 lS87.2 272.0 1128.9 654.9 1374.7 405.0 1886.7 1744.0 494.0 2266.1 1953 2850.9 558.1 346. 7. a. station (8) Compute the frequency curve using adjusted values of y and S in Equation 3-4 and K values from Appendix F-2 corresponding to the adopted skew coefficient.3 ----------------------------------------------------------------------------------------------- Note . Nature and Purpose.3 742.7 423.9 llos.s 972.0 1216.7 1152.3 476.1 732. 3.7 930.0 1936.9 1145. and the confidence limits.0 608.9 632.9 380. year through September 30 3-18 .7 1963 1730.5 530.5 515.6 1186. 120.2 1946 1360. Runoff volumes are expressed as average flows in order that peak flows and volumes can be readily compared and coordinated.7 13S8.0 1117.7 649. Flood volume frequency studies involve frequency analysis of maximum runoff within each of a set of specified durations.1 794.3 1644.0 566.2 669.EM 1110-2-1415 5 Mar 93 (7) Calculate the equivalent length of record of the mean for the short-record using Equation 3-20.7 462.2 379.0 375.3 581.8 1572.0 1916.0 2070.Data based on water year of October I of preceeding of given year.0 2966.

but should be weighted with information from regional studies. the longer-duration flood volumes for that year can often be estimated adequately. the skew coefficient should not be based solely on the station record. only the first two moments. care should be exercised to assure that the period selected is the one when flows would have been maximum under the specified (usually natural) conditions. Each of these values can be found by substitution into Equation 3-4 (the K for zero skew and 99. + (-3..719) sa Za .269. Data to be used for a comprehensive flood volume-duration frequency study should be selected from nearly complete water year records. there is a maximum and minimum allowable slope for the standard deviation-versus-mean relation which prevents the curves from crossing within the established limits. Data for Comorehensive Series. For instance.TT. the durations in Table 3-2 should be used in order to assure consistency among various studies for comparison purposes.EM 1110-2-1415 5 Mar 93 b. 2 0.99 percent chance exceedance Ti. The value of this slope constraint is found by stating that the value of one curve (XA for curve A) must equal or exceed the value for a second curve (Xa for curve B) at the desired exceedance frequency. In practice. As discussed in Section 3-3. 3.719): x.719) s.269 where: X* = Value of frequency curve A at 99. nor be less than -. and especially to prevent the curves from crossing.719 (Se . Unless overriding reasons exist.99 percent chance exceedance is -3. mean and standard deviations are based on station data. (I) The probability distribution recommended for flood volume-duration frequency computations is the log-Pearson type III distribution. L x..99 percent chance exceedance Xr3 = Value of frequency curve B at 99. For a given skew coefficient. = Mean of frequency curve B % = Standard deviation of frequency curve A sil = Standard deviation of frequency curve B 3-19 . the same as that used for annual flood peaks. = Mean of frequency curve A Yi.q z 2 TI. Where a minor portion of a water year’s record is missing. the slope must not exceed . to keep the curves from crossing within 99. c. + (-3.) (ST.99 and . Maximum flood events should be selected only for those years when recorder gages existed or when the maximum events can be estimated by other means. This can be done graphically as shown in Figure 3-6. If upstream regulation or diversion is known to have an effect. respectively. Statistics for Comorehensive Series.269.) (S* .Ol percent chance exceedances with a skew of O. it is desirable to coordinate the variation in standard deviation and skew with that of the mean. To insure that the frequency curves for each duration are consistent.S.S. Tz.

.10 2.EM 1110-2-1415 5 Mar 93 0.15 0. Coordination 3-20 of Flood-Volume Statistics.6 - 2.2 3.26 4 B 5 z 0.0 3.W ci 0.4 2.6 2.4 2. 0. Figure 3-6.4 Mean of Lo(laibhm (Z.2 MesIn of Logartthns (X.0 3.6 2.6 3.6 3.20 ci 3 5 0.

. This is a general check on the analytic work and will ordinarily reveal any inconsistency in data and methodology. The procedure for computing a least-squares line through a given intersection can be found in texts describing regression analyses. the smoothing relations should be forced through those points. Freauencv Curves for ComDrehensive Series. it is probably easiest to adopt smoothed relations for the standard deviation and skew and input the statistics into a computer program that computes the ordinates. (1) Ge e al p ocedu e Frequency curves of flood volumes are computed analytically :si’ng g&era1 brinciples and methods of Chapters 2 and 3. Figure 3-7. (3) If the statistics for the peak flows have been computed according to the procedures in Bulletin l7B. 07 L 0 z M c u 5 n E LL z A IA z I2 I Figure 3-7. They should also be shown graphically and compared with the data on which they are based.EM 1110-2-1415 5 Mar 93 (2) When the skew changes between durations. The curves can then be inspected for consistency. Flood-Volume 3-21 Frequency Curves. d. The computed frequency curves and the observed data should be plotted on a single sheet for comparison purposes.

e. (2) Abblication to a Single Reservoir. In the case of a single flood-control reservoir located immediately upstream of a single damage center. Flood control reservoirs are designed to substantially affect the frequency of flood flows (or flood stages) at various downstream locations. These are drawn first on logarithmic paper for interpolation purposes. as illustrated on Figure 3-8a. and operational contingencies. can also have significant effects on downstream flood flows (see Section 3-10). forest clearing. design. encroachment into the flood plain. etc. (3) ADDlication to a Reservoir Svstem. special frequency studies should be made. The degree to which flows and stages are modified by various flood control works or land use changes can depend on the timing.000 cfs-days (31. the studies should include evaluations of the effects on representative flood 3-22 . (1) Volume-duration Curves. Levees can also create backwater conditions that affect river stages for a considerable distance upstream. Nature. Many land use changes such as urbanization.500 cfs.EM 1110-2-1415 5 Mar 93 (2) c Durations. The runoff volume for any specified frequency can be determined for any duration between l-day and 365-days by drawing a curve on logarithmic paper relating mean discharge (or volume) to duration for that specified frequency (see Figure 3-&la). but also as an aid in drawing filling-frequency curves. and operation problems usually involves the construction of volume-duration curves for specified frequencies. representative hydrographs at all locations can be patterned after one or more past floods. The curve also indicates that a duration of about 8 days is critical for this project release rate and associated flood-control storage space. similar to that shown on Figure 3-8. the volume frequency problems are relatively simple. This procedure can be used not only as an approximate aid in selecting a reservoir capacity. ADDliCatiOnS of Flood Volume-Duration Freauencies. These could be done in the same manner as for the longer durations. A design or operation scheme based on regulation of such a set of hydrographs would be reasonably well balanced. The ordinates of these hydrographs can be adjusted so that their volumes for the critical durations will equal corresponding magnitudes at each location for the selected frequency. corresponding to selected exceedance frequencies should first be drawn.500 cfs. The straight line represents a uniform flow of 1. areal distribution and magnitude of rainfall (and snowmelt. A straight line on this grid represents a constant rate of flow. if pertinent) causing the flood. The use of flood volume-duration frequencies in solving reservoir planning. Some aspects of this problem are described in Section 3-9g. a. A series of volume-duration curves. When runoff volumes for durations shorter than 24 hours are important. Channel improvements (intended to reduce stages) and levee improvements (intended to confine flows) at specified locations can substantially affect downstream flows by eliminating some of the natural storage effects. and the maximum departure from the 2% chance exceedance curve demonstrates that a reservoir capacity of 16. 3-9. In solving complex reservoir problems. Accordingly. giving due consideration to possible channel deterioration. Effects of Flood Control Works on Flood Freauencies. The project release rate should be determined. using skew coefficients interpolated in some reasonable manner between those used for peak and l-day flows.700 acre-feet) is required to control the indicated runoff volumes by a constant release of 1. The mean discharge values are multiplied by appropriate durations to obtain volumes and are then replotted on an arithmetic grid as shown on the Figure 3-8b.

MEAN FLOW.I* . F 2 w b 01* 0 . z: m i.VOLUME.: .-l-L I t 1-I - z 3 e. e..: -z: -:: Fi mm: “. lo =I I z g P E g’ (n 5 i I 2 ? oc E 3 0 n a P fi z Y _ b-... z x t 4. -it z. orI : . In* _n .g 0 .A: xi: 001 : w I g w . 2 t . CFS CFS-DAYS n L.-.L. -4 -a-. --. E E 7 I I I 1 !z “7 SCM I :.P awul .

EM 1110-2-1415 5 Mar 93

events, with careful consideration distributions. b. Terminology.

given to the effects of different

temporal and area1

(1) Natural Conditions. Natural conditions in the drainage basin are defined as hydrologic conditions that would prevail if no regulatory works or other works of man were constructed. Natural conditions, however, include the effects of natural lakes, swamp areas, etc. (2) Present Conditions. Present or base conditions are defined as the conditions that exist as of the date of the study or some specified time. (3) Without-Proiect Conditions. Without-project conditions are defined as the conditions that would exist without the projects under consideration, but with all existing projects and may include future projects whose construction is imminent. (4) With-Proiect Conditions. With-project conditions are defined as the conditions that will exist after the projects under consideration are completed. c. J? rv irQa.Level+reauencv. (1) Factors to be Considered. Factors affecting the frequency of reservoir levels include historical inflow rates and anticipated future inflow rates estimated by volume-frequency studies, the storage-elevation curves, and the plan of reservoir regulation including location and size of reservoir outlets and spillway. A true frequency curve of annual maxima or minima can only be computed when the reservoir completely fills every year. Otherwise, the events would not be independent. If there is dependence between annual events, the ordinate should be labeled “percent of years exceeded” for maximum events and “percent of years not exceeded” for minimum events. (2) Combutation and Presentation of Results. A frequency curve of annual maximum reservoir elevations (or stages) is ordinarily constructed graphically, using procedures outlined in Section 2-4. Observed elevations (or stages) are used to the extent that these are available, if the reservoir operation will remain the same in the future. Historical and/or large hypothetical floods may also be routed through the reservoir using future operating plans. A typical frequency curve is illustrated on Figure 6-4. Elevation-duration curves are constructed from historical operation data or from routings of historical runoff in accordance with procedures discussed in Section 2-2, Figure 3-9. Such curves may be constructed for the entire period of record or for a selected wet period or dry period. For many purposes, particularly recreation uses, the seasonal variation of reservoir elevation (stages) is important. In this case a set of frequency or duration curves for each month of the year may be valuable. One format for presenting this information is illustrated on Figure 3-10.

3-24

**EM 1110-2-1415 5 Mar 93
**

1010

EXAMPLE

RESERVOIR

-TOP OF FIDOD CONTROL moL

L L

i :: 2 z IO20id

lOi0

1030 i-

_f

..-

-_-___

A

.+..-.

i

'

i

1

i lolo.:

-fo? OF VAT6R su?rLY PWL

1000 m.0 n.a as

P&NT

OF fIWWE EXCEEC&

1

.I

.Ol 976.0

I 1

-m?

OF

INACIIVE

Figure 3-9. Daily Reservoir Elevation-Duration

Curve.

PorcuM of Oaya Exceeded

Figure 3- 10. Seasonal Variation of Elevation-Duration 3-25

Relations.

EM 1110-2-1415 5 Mar 93 d. Fffects of Reservoirs on Flows at Downstream Points. (1) Routine for Period of Record. The frequency of reservoir outflows or of flows at a downstream location can be obtained from routings of the period-of-record runoff by the following methods: (a) Determine the annual maximum flow at each location of interest and construct a frequency curve of the regulated flows by graphical techniques (Section 2-4). (b) Construct a graph of with-project versus without-project flows at the location of interest and draw a curve relating the two quantities as illustrated on Figure 3-l 1. The points should be balanced in the direction transverse to the curve, but factors such as flood volume of the events and reliability of regulation must be considered in drawing the curve. This curve can be used in conjunction with a frequency curve of without-project flows to construct a frequency curve of with-project flows as illustrated on Figure 3-12. This latter procedure assures consistency in the analysis and gives a graphical presentation of the variability of the regulated events for a given unregulated flow. (2) Use of Hvoothetical-Flood Routinns Usually recorded values of flows are not large enough to define the upper end of the regulated frequency curve. In such cases, it is usually possible to use one or more large hypothetical floods (whose frequency can be estimated from the frequency curve of unregulated flows) to establish the corresponding magnitude of regulated flows. These floods can be multiples of the largest observed floods or of floods computed from rainfall; but it is best not to multiply any one flood by a factor greater than two or three. The floods are best selected or adjusted to represent about equal severity in terms of runoff frequency of peak and volumes for various durations. The routings should be made under reasonably conservative assumptions as to initial reservoir stages. (3) Incidental Control bv Water SUDD~V Soace. In constructing fre:uency curves of regulated flows, it must be recognized that reservoir operation for purr .es other than flood control will frequently provide incidental regulation of floods. Ii .wever, the availability of such space cannot usually be depended upon, and its value is considerably diminished for this reason. Consequently, the effects of such space on the reduction of floods should be estimated very conservatively. (4) da. All w n r In constructing frequency curves of regulated flows, it should be recognized that actual operation is rarely perfect and that releases will frequently be curtailed or diminished because of unforeseen operation contingencies. Also, where flood forecasts are involved in the reservoir operation, it must be recognized that these are subject to considerable uncertainty and that some allowance for uncertainty will be made during operation. In accounting for these factors, it will be found that the actual control of floods is somewhat less than could be expected if full release capacities and downstream channel capacities were utilized efficiently and if all forecasts were exact. (5) Runoff from Unreaulated Area& In estimating the frequency of runoff at a location that is a considerable distance downstream from one or more reservoir projects, it must be recognized that none of the runoff from the intermediate areas between the reservoir(s) and the damage center will be regulated. This factor can be accounted for by constructing a frequency curve of the runoff from the intermediate area, and using this curve as an indicator of the lower limit for the curve of regulated flows. Streamflow 3-26

10 ‘s ‘7 10 hmulat-o cfs Figure 3-11. 1o4S9.mmt Chmcm E xmwd-m 60 10 1 .il Figure 3-12. Example Without-Project 3-27 and With-Project Frequency Curves. .1 .9 S“s 9“9 P. Example With-Project versus Without-Project Peak F1OWRelations.EM 1110--2-1415 5 Mar 93 10 10 10 10 1 104 10 ‘s P--k Flou. SS 99.

f. ~ han in ionsh~. Urbanization has two major effects on the watershed which a. g. (2) ComDute r Pr~ra m . i. It is generally impossible to make all of the flood routings necessary to evaluate the effect of a reservoir system by hand computations. snowmelt. The resulting relationships for modified conditions can be used to modify rotlting criteria to enable evaluation of the downstream effects of these changes. Effects of Mu ltiDle Rese rvoir Svste~. it is best to make complete routings of five to ten historic flood events and a large event that has been developed from a hypothetical rainfall pattern. there should be a balance of events caused by particular climatic factors. First.project values.e. Care muse be exercised in selecting events that have representative flood volumes. and a smooth curve drawn. Changes in stage-discharge relations due to channel improvements. which results in more water entering the stream system as direct runoff. The stages or discharges thus derived can be plotted against corresponding without. Corresponding stages upstream from the selected control point can be estimated from water-surface profile computations. Hypothetical events must be used with caution. however. etc. If necessary. Second. Also. Whenever practicable. thunderstorm. the flood-volume-duration characteristics of the hypothetical events should be similar to the recorded events (see Section 3-8). Furthermore. the drainage system collecting the runoff is generally more efficient and tends to concentrate the water faster in the downstream portion of the channel system. Effects of Channel. such studies should be supplemented by a critical examination of the potential effects of atypical floods. Accordingly. Ieveee-and floodway improvem~nts on river stages at the project location and on river discharges downstream from the project location can generally be evaluated by routing several typical floods through the reaches of the improvement and the upstream reaches affected by backwater. (I) Ret)resentative events. the problem of evaluating reservoir stages and downstream flows under project conditions becomes increasingly complex. Effects of Urbanization. When more than one reservoir exists above a damage center. ~ influence the runoff characteristics. 3-28 . The effect of channel. Computer programs have been developed to route floods through a reservoir system with complex operational criteria (55). 3-10. timings. Le ee and Floodwav ImDrovemen&. because certain characteristics of atypical floods may be responsible for critical flooding conditions. This curve could be used in conjunction with a frequency curve of without-project values to construct a frequency curve of with-project values as discussed in Paragraph 3-09d( 1)b. levee construction or flow obstructions can best be evaluated by computing theoretical water surface profiles for each of a number of discharges.EM 1110-2-1415 5 Mar 93 routing and combining of both the flows from the unregulated area and those from the regulated area is the best procedure for deriving the regulated frequency curve. It is important to keep these two effects in mind when considering the changes in the flood peak frequency curve caused by increasing urbanization. tropical storm. it is possible to supplement these events by using multiples of the flow values. there is a substantial increase in the impervious area. and areal distributions.

Typical Effect of Urbanization on Flood Frequency Curves. Effect on Freauencv Relations.999s. thunderstorm.lt is possible for urbanization to cause a decrease in the flood peaks at a particular site. This change in the timing of the peaks would result in lower downstream peaks. c. basin shape. Graphical techniques should be applied if a good fit is not possible by an analytical distribution.13). when both areas have become equally urbanized.01 Figure 3-13. A general statement can be made about the effects of urbanization on flood-peak frequency relations. hurricane. consider an area downstream of two tributary areas of such size and shape that the large floods are caused by the addition of the nearly coincident peaks from the two tributaries. Other Considerations. but causes backwater flooding upstream. The construction of bridges or other encroachments can reduce the flood peak downstream. the flood peaks may coincide again. Some of the factors that must be considered are basin slope.1 . For instance. Of course.9 99 90 Percent 50 Chncm Ex~=tince 10 1 . but a lesser increase in the less frequent events. Urbanization in one of the tributary areas will likely cause the contribution from this area to arrive downstream earlier. number of depressional areas drained. 99. 3-29 .EM 1110-2-1415 5 Mar 93 b. or frontal storm). The actual effect of urbanization at a specific location is dependent on many factors. magnitude and nature of urban development and the typical flood source (snowmelt. a decrease in the standard deviation and an unpredictable effect on the skew coefficient (see Figure 3. The resulting frequency relation may not fit any of the standard theoretical distributions. This results in an increase in the mean of the annual flood peaks. The usual effect on the frequency relation is to cause a significant increase in the magnitude of the more frequent events. previous land use and ground cover.

EM 1110-2-1415 5 Mar 93 d. Adius tment of a Series of Nonstationary Peak Discharges. When the annual peak discharges have been recorded at the outlet of a basin which has been undergoing progressive urbanization during the period of record, the peak discharges are nonstationary because of the varying basin condition. It is generally necessary to adjust the discharges to a stationary series representative of existing conditions. One approach to adjusting the peaks to a stationary series is as follows (1) Develop and calibrate a rainfall-runoff model for existing basin conditions and for conditions at several other points in time during the period of record. (2) Develop a hypothetical storm for the basin using generalized rainfall criteria, such as that contained in Weather Bureau Technical Paper 40 ( 14). Select the magnitude of the storm, e.g., a 25-year recurrence interval, to be used. The recurrence interval is arbitrary as it is not assumed in this approach that runoff frequency is equal to rainfall frequency. The purpose of adopting a specific magnitude is to establish a base storm to which ratios can be applied for subsequent steps in the analysis. (3) Apply several ratios (say 5 to 8) to the hypothetical storm developed in the previous step such that the resulting calculated peak discharges at the gage will cover the range desired for frequency analysis. Input the balanced storms to the rainfall-runoff model for each of the basin conditions selected in step ( 1), and determine peak discharges at the gaged location. (4) From the results of step (3), plot curves representing peak discharge versus storm ratio for each basin condition (or point in time). (5) Use the curves developed in step (4) to adjust the observed annual peak discharges. For example, an observed annual peak discharge that occurred in 1975 is adjusted by entering the “1975” curve (or interpolating) with that discharge, locating the frequency of that event, and reading the magnitude of the adjusted peak from the base-condition curve for the same frequency. The adjusted peak thus obtained is assumed to be the peak discharge that would have occurred for the catchment area and development at the base condition. It is not necessary to adjust to natural conditions. A stationary series could be developed for one or more points in time. (6) A conventional frequency analysis can be performed on the adjusted peak discharges determined in the preceding step. If the data represent natural conditions, Bulletin 17B procedures would be applicable. If the basin conditions represent significant urbanization, graphical analysis may be appropriate. e. Development of Freauencv Cu rves at Unea~ed Slte$. There are several approaches that can be taken to develop frequency curves at ungaged sites that have been subject to urbanization. In order of increasing difficulty, they are 1) application of simple transfer procedures (e.g., Q = CIA); 2) application of available region-specific criteria, e.g., USGS regression equations; 3) application of rainfall-runoff models to hypothetical storm events; 4) application of simple and detailed rainfall-runoff models with observed storm events and 5) complete period-of-record simulation. As approaches (3) and (4) are often applied, the computational steps are presented in some detail.

3-30

EM 1110-2-1415 5 Mar 93 (1) Hypothetical Storm Approach (a) Develop peak-discharge frequency curve for specific land use conditions from available gaged data and/or regional relationships. (b) Develop balanced storms of various frequencies using data from generalized criteria, a nearby gage or the equivalent. (c) Develop rainfall-runoff model for the specific watershed with the adopted land-use conditions. Calibrate runoff and routing parameters by reproducing observed hydrography occurring under natural conditions. (d) Input balanced storms (from b) to rainfall-runoff model (from c). Determine exceedance probabilities to associate with balanced storms from adopted specific land-use conditions peak discharge frequency curve (from a) with computed peak discharges. (e) Modify parameters of rainfall-runoff model to reflect future urban runoff characteristics. Input balanced storms to the urban- conditions model. (f) Plot results assuming frequency of each event is the same for both the adopted land use and the future urban conditions. (2) “Simple” and Detailed Simulation of Historic Events (a) Simulate all major historic events with a relatively simple model to establish the ranking of events and an approximate peak discharge for each. The approximate peaks could be developed by using a multiple linear regression approach, by using a very simple rainfall-runoff model, or by any other approach that will capture the hydrologic response of the basin. (b) Perform a conventional frequency analysis of the approximate peaks obtained in step a. (c) Make detailed simulations of selected events and correlate the more precise peaks with the approximate peaks. (d) Use the relationship developed in step c to determine the desired frequency curve. The same approach can be followed for both existing and future conditions.

3-31

EM 1110-2-1415 5 Mar 93 CHAPTER4 LOW-FLOWFREQUENCY ANALYSIS

4-1. ~. Low-flow frequency analyses are used to evaluate the ability of a stream to meet specified flow requirements at a particular location. The analysis can provide an indication of the adequacy of the natural flow to meet a given demand with a stated probability of experiencing a shortage. Additional analyses can indicate the amount of storage that would be required to meet a given demand, again with a stated probability of being deficient. The design of hydroelectric power plants, determination of minimum flow requirements for water quality and/or fish and wildlife, and design of water storage projects can benefit from low-flow frequency analysis. 4-2. InterDretat ion. a. Analytical frequency techniques are usually not applicable to low-flow data because most theoretical frequency distributions cannot satisfactorily fit the recorded data. It is recommended that graphical techniques be used and that known geologic and hydrologic conditions be kept in mind when developing the relationships. As the low values are the major interest, the data are arranged with the smallest value first. The probability scale is usually labeled “percent chance nonexceedance.” b. Annual low flows are usually computed for several durations (in days) with the flow rate expressed as the mean flow for the period. For example, the USGS WATSTORE output provides the mean flow values for daily durations of 1, 3, 7, 14, 30, 60, 90, 120 and 183 days. The default values for the HEC program STATS are the same with the exception of using a 15-day duration instead of 14 (Table 4-1). Often a climatic year from April 1 to March 31 is specified to provide a definite separation of the seasonal low-flow periods. Figure 4-1 is a plot of the data in Table 4-1. 4-3. ADDlication Problem%.

a. Pas in Develonmen~. The effec~ of any basin developments on low flows are usually quite significant. For example, a relatively moderate diversion can be neglected when evaluating flood flow relations, but it would reduce, or even eliminate, low flows. Accordingly, one of the most important aspects of low flows concerns the evaluation of past and future effects of basin developments. b. Multi-Year Events. In regions of water scarcity and where a high degree of development has been attained, projects that entail carryover of water for several years are often planned. In such projects it is desirable to analyze low-flow volume frequencies for periods ranging from 1- 1/2 to 8-1/2 years or more. Because the number of independent low-flow periods of these lengths, in even the longest historical records, is very small and because the concept of multi- annual periods is somewhat inconsistent with the basic concept of an “annual event;” there is no truly satisfactory way for computing the percent chance nonexceedance for low-flow periods that are more than 1 year in length. One procedure described in reference (37) has been used with long sequences of synthetically generated streamflows to derive estimates of drought frequency. Although

4-1

2 5. 4-2 .9 19.4 21.----------------------------. it is impossible to verify the accuracy of the frequency estimates.3 1955 7. NY .6 48.6 119.7 254.2 7. -------------------------------- YsAR CFS VALUE FOS DURATION. etc.8 60.2 67.6 14.0 6.6 78.0 10.7 15.0 45.3 9.9 22.2 11.0 51.6 14.0 4.2 11.1 6.8 1954 8.2 15. It may be possible to estimate low-flow rates on a per unit area basis for a given exceedance frequency if the study area is relatively homogeneous with respect to geology.9 32.7 179.3 53.8 98.5 9.2 31.1 27.4 2::: 7.5 36.9 32.0 104.4 1965 5.1 10.6 50.3 11.4 S2.5 1966 1967 43.4 12..-----------------.3 44.9 4.3 122.3 10.1 1956 17. Re~ionalization.FISEKILL CR AT BSACON .4 58..3 20.3 1952 39.2 41. are not easily quantifiable to enable translation into probable low-flow rates.--------------------------.0 13.3 12.EM II1O-2 1415 5 Mar93 Table 4-1.0 62.1 27.1 13. The variations in geologic conditions such as depth to ground water. Low-Flow Volume-Duration .5 26.7 5.2 57.7 34. o 1953 4.7 6.4 44. s 20.0 ---------------------------.0 97.. Regionalization of low-flow events is usually not very successful.8 1.5 29.6 2::: 118.9 4.8 1. and climate.3 75.6 135.1 77.0 33.0 19.0 h3.7 1962 7.0 5::.0 21.9 15.5 29.VOLW-DURATION DATA .1 1946 9.2 .3 1060 53.0 111.0 1950 22. size of ground water basin. the procedures described by Riggs (28) should be reviewed for applicability.0 19.6 1964 1.---------------1945 92.9 20.9 41.4 62.0 19.3 23.------------220.1 8.7 84.3 1948 8.1 53.1 64.9 14. permeability of the aquifer.1 1.8 17.5 20.4 12.2 43.2 11.3 1961 17.4 25.3 49.3 40.4 9.8 6.1 46.2 39. topography.3 1959 17.6 5.5 12.--------note Data bud on Climtic Yeu W Data.5 143.9 1963 19.7 160.1 37.0 32.6 15.1 21.4 4.-----.DAILY F~ -----. 122.2 137.9 18.9 22. F-.8 13.6 1947 0.0 73.1 9.0 6.3 100.6 16.0 17.0 9.1 305.0 49.-.2 59.0 15.4 99.3 1956 19.7 4.0 108.0 1957 3.5 21.0 127.-----------------------------------------L~ST 1 3 7 -------------------.4 37. 26.4 17.6 25.5 27. 15 ------- the results obtained through the use of this procedure seem reasonable.:.0 22.9 149.8 62.1 32.4 116.2 9.9 136.4 21.5 49.0 23.0 21.7 247.0 115.3 25.7 1949 7.0 19.9 64.2 88.1 28.8 140.6 147.6 5.0 58.6 12.6 39.0 17.af next year.7 33.0 63.9 70.4 12.7 12.0 68.0 4.7 41.5 15.----------.7 9.5 58.1 19. 30 60 90 120 183 -.9 71.8 6. c.1 10.5 213.1 37. of April 1 of given you through March 31 .0 27.0 1951 40. If information is needed at several ungaged sites.

EM 1110-2-1415 5 Mar 93 ~03 FIS} EK AT BEACON.~2_ _ lo—. Low-Flow Frequency Curves.1 — . 4-3 .01 PERCENT CHANCE NONEXCEEOANCE Figure 4-1. NY 1945-1960 k .9 9s 90 so 10 1 .99 69.- IF l— \ 7-DAY 3-OAY — 99.

instantaneous peak intensities are ordinarily not analyzed since they are virtually impossible to measure and are of little practical value. The logarithmic normal. and some method of evaluating the frequency of simultaneous or near-simultaneous precipitation over an area is necessary. ui= WSU. mostly for durations of less than 3 or 4 days.1 . Both graphical and analytical methods may be used. Procedures for obtaining depth-area frequency curves are usually available from National Weather Service publications (references are given in subsequent paragraphs). The National Weather Service has traditionally used the Fisher-Tippett Type I frequency distribution with Gumbel’s fitting procedure.01 Figure 5-1. Station precipitation alone is not adequate for most hydrologic studies. Annual Precipitation.EM 1110-2-1415 5 Mar 93 CHAPTER5 PRECIPITATION FREQUENCY ANALYSIS 5-1. pearson Type III and log-Pearson Type III (Figure 5-1) distributions. Cumulative precipitation amounts for specified durations are commonly analyzed. however. In precipitation studies. Ge neral Proced es The computation of frequency curves of station precipitation can be ~~ne”by procedures similar to those for streamflow analysis described in the preceding sections. 3s 90 Pmrc=nt Cknce se ExceeMnce 10 1 . have also been used with success. 5-1 . . Frequency Curve.

It must be remembered that a frequency curve computed from observed flood peaks is based on a relatively small sample.frequency-duration studies that have incorporated regional information. This results because not all the possible loss rates for a given magnitude of precipitation are modeled. The resulting flood frequency relations must be adjusted if an annual peak flood frequency relationship is desired. Longer duration events (2. Guidelines for developing runoff frequencies from precipitation frequencies are presented in references ( 10) and (44). Precipitation-frequency flood:frequency relations where inadequate flow data are available or where existing (or proposed) watershed changes have modified (or will modify) the rainfall-runoff relationships. Flood-frequency curves developed by rainfall-runoff procedures often have less variance (lower standard deviation) than those developed from annual flood peaks. an appropriate ratio adjustment for the standard deviation should be developed for the region. the derivation of frequency relations by rainfall-runoff modeling requires careful checking for consistency at every step. It is possible that the flow-frequency curve derived from precipitation-frequency data is more representative of the population flow-frequency curve than the one computed from the statistics of the observed flood peaks. Reference (44) describes the procedures involved in calibrating a HEC. Calibration.1 model to a flow-frequency curve based either on gaged data from a portion of the basin or on regional flood-frequency relations. These reports have maps for 6. Or. The coefficients from the calibrated model must be consistent with those from nearby basins that have also been modeled. If extensive use will be made of frequency curves derived by rainfall-runoff modeling. For durations of 5 to 60 minutes in an area generally east of 105th meridian. these relations are from a partial-duration series. Where practical.and 24-hour durations with extrapolation procedures to obtain durations less than 6 hours. ~ial relations presented in the National Weather Service publica~ions represent all the events above a given magnitude. Duration The precipitation-frequency c. Available~onal Inform ation. (22) and (23). individual reports have been prepared for each of the 11 western states (24). Derivation of Flood-Freauenc v Relations from Prec iDitat ion.to 10-days) are presented in references (21). more typically. use should be made of previous precipitation. Therefore. For durations of 2 to 24 hours in the same area see Technical Paper 40 (10). But.EM II1O-2-1415 5 Mar 93 5-2. 5-3. Because of the orographic effect. there are also errors in calibrating the model and establishing loss rates approximate with the different frequency events. relations are often used to derive ADDIicat ion. the partial-duration series precipitation estimates are adjusted to represent annual series estimates prior to use. therefore. 5-2 . b. see Hydro-35 (9).

6-1 .) 6-2.EM 1110-2-1415 5 Mar 93 CHAPTER6 STAGE (ELEVATION) . some sites located near estuaries have only stage information because the flow is affected by varying backwater conditions. in addition to annual peak flows. Flow Frequency Curve. -. Minimum elevation-frequency curves are used to evaluate the recreation benefits at a lake or reservoir. staee Data.“ . Also.FREQUENCY ANALYSIS 6-1. maximum annual stages at most sites. b. s Unre@sted Computed Frequency Curve Derived Curve by Graphlcel Procedures Regulated Frequency Ss so P9mmt se 10 -cm I?xw-dem 1 .“ . elevations are generally referenced to mean sea level. storm surges along a lake or ocean.“ . to evaluate minimum depths available for navigation purposes. wind driven waves (runUp). Inundation can result from a flooding river.“ . the construction of levees. bridges. etc. to locate a water supply intake. whereas. -------I — 99. or combinations of any of these. a. Maximum stage-frequency relations are often required to evaluate inundation damage. River stages can be very sensitive to changes in the river channel and floodway.01 Figure 6-1. The USGS WATSTORE Peak Flow File has.99 Os. Unregulated and Regulated Conditions. Therefore. (Stages are referenced to an arbitrary datum. or channel modifications can result in stage data that is non-homogeneous with respect to time.“ .. For riverine situations.“1 .“ : 11 7 4 . it is usually recommended that the flow-frequency curve (Figure 6-1 ) and a rating curve (stage versus . ~. a filling reservoir.

Historical information can be incorporated into a graphical analysis of stage (elevation) data by use of the procedures in Appendix 6 of Bulletin 17B (ref 46). top of conservation pool. a downstream reservoir. Usually the annual extreme value is used to develop an annual series. 6-3. spillway elevation. c. Independent events are not easily determined if the events are elevations of a large lake or reservoi~ in fact even the annual events may be significantly correlated. A stage-frequency curve derived by indirect methods may not always represent the true relation if the site is subject to occasional backwater situations. known constraints must be kept in mind. Caution must be used in selecting independent events. Stage (elevation) data are usually not normally distributed (not a straight line on probability paper). a debris flow. an analytical analysis should ~ be made without observing the fit to the plotted points (see Chapter 2). a storm surge. Therefore. bankfull stage. top of flood control pool. Figure 6-2) beusedto derive astage-frequency curve (Figure 6-3). Usually. The expected probability adjustment should be made to the flow-frequency curve when the stage (elevation) frequency relation is derived indirectly. all influence the elevation-frequency relation for a reservoir. As an example. would be important for a riverine site. The median plotting position formula corrects for the bias caused by small sample sizes. but there may be situations where a logarithmic or some other appropriate transformation will make the plot more nearly linear. Again. When drawing the curve. or a downstream river. etc. Figure 6-4. These constraints usually make these frequency relations very non-linear. Freaue ncv Distribution. 6-2 . etc.EM 1110-2-1415 5 Mar 93 flow. The statistical tests (Appendix 4.. but a seasonal series or a partial-duration series could be developed if needed. Extrapolation of stage (elevation) frequency relations must be done very cautiously. any constraints acting on the relations must be used as a guide in drawing the curves. a high tide. Backwater conditions can be caused by an ice jam. The minimum pool. ref 46) to screen for outliers should N be applied unless the stage (elevation) data can be shown to nearly fit a normal distribution. operation criteria. A coincident frequency analysis may be necessary to obtain an accurate estimate of the stage-frequency relationship (see Chapter 11). The expected probability adjustment should be made when an analytical method is used directly to derive the stage (elevation) frequency relation. levee heights. Exnected Proba bilitv. the bottom elevation. an arithmetic-probability plot is appropriate for stage or elevation data. The expected probability adjustment should ~ be made to frequency relations derived by graphical methods. 6-4..

4 . 1949-8S Nov 12. . . .“ Derived Stage-Frequency Derived St~e-Frequency I i 6S Curve. . 3a q . ----. . .s1 1 so Pmmmnt l-a Figure 6-3. .EM 1110-2-1415 5 Mar 93 Obaemad Peak F1OWSand Stages. .. Unregulated and Regulated Conditions 6-3 . S Curve.“ . Rating Curve for Present Conditions. Regulated I 60 ChMm E xrnmd~= Conditions 1 I 1 . . i . ..1 . . . Derived Stage-Frequency Curves. . . — Rating Curve for Gage Locstion. 1948 to Praaent e seooe zeeooa 160800 aoeoso a6eo@e 3oeeeo Figure 6-2. 26 .. . . . . . Unregulated Conditions — la SS: = S6-.

1 . S--TO? OF OAN ! toso 1 ! lu7.01 978.. SS i S9. 6-4 . Maximum Reservoir Elevation-Frequency Curve.ANNUAL MAXIMUM ELEVATIONS I ~ 10S6 .5- - TOP OF INDUC~ SURCliARCE loA6.E INA~Ivs Figure 6-4.EM 1110-2-1415 5 Mar 93 i 060 CXAMPLF mtSERVO19 .9 2s PER:ENTCHANC.-TOP OF WD CONTROL PoOL I 1040 to30 I 1020 I J d 1010.0 L TOP OF EXCEEDA.0. 1 .-TOP OF UATSR SUPPLY POOL tolo 1000 S9.5.

historic damage could be scaled to the present to account for price differences (inflation) and the average simply computed.EM 1110-2-1415 5 Mar 93 CHAPTER7 DAMAGE-FREQUENCY 7-1. If 50 years of damage information were available for an area that has remained in essentially the same land use with a reasonably constant level of economic activity. Care should be taken to assure the rating curve is not looped so that discharge is a unique function of stage. This technique is described in detail in Section 7-2. and discharge-exceedance frequency functions) for each index location for existing conditions. a. a. Introduction. Sensitivity analysis may be useful in determining the reliability of the computed expected annual damage considering the uncertainties involved. There are three methods that may be used to compute average annual damage and are herein termed the historic method. The steps involved in determining the reduction in annual damage due to project measures are: (1) Develop the basic relationships (stage-damage. This technique addresses the disadvantages of the previous two methods. The average of the time trace of damage would be the average annual damage. The most widely used approach within the Corps of Engineers is the frequency technique. b. c. Commutation of ExDected Annual Dama~e. or the historic record used. the simulation method and the frequency method. stage-discharge. Otherwise more complex functions that correctly relate stage and discharge should be developed and applied. RELATIONSHIPS 7-1 . along with damage functions to generate a time trace of simulated damage. Figure 7-1 shows a schematic of the application of the three basic damage evaluation functions used to compute the expected value of the annual damage. b. The simulation method has the advantage of permitting the use of complex damage functions that can consider more than a single parameter and thus enable a more accurate computation of damage. Experience in the development and application of damage functions is essential to computation of reasonable estimates. 7-2. Damage functions in agricultural areas are often a function of the season and the duration of flooding. A hydrologic simulation model could be developed. This approach is termed the historic method and is the most direct but is seldom used because sufficient data usually do not exist and the land use and economic activity of an area are usually changing. This would be termed the “simulation” method. and yet is fairly easily applied. The term “expected” is used rather than “average” because a frequency curve is used to represent the distribution of future flood events and the expected value of damage is computed by the summation of probability weighted estimates of damage. The disadvantage of this method is that the future floods are assumed to exactly duplicate the historic floods and no consideration given to the possibility of larger floods.

Cfs Dmga Relatii [ Figure 7-1. Schematic for Computation of Expected Annual Damage. 7-2 .EM 1110-2-1415 5 Mar 93 Discharge-Fr~y Cuve Rati me ~~ Discharge.

b. there may be long-term adjustments in alluvial channel flow regimes that would cause the rating curve to change with time. a. (5) Repeat step (1) for each alternative flood plain management plan under investigation. revise the three basic evaluation functions as necessary. Watershed runoff characteristics may be changing with time due to changes in land use.e. 7-3. The computer program “Expected Annual Flood Damage Computation” (54) has the capability to make these computations. (6) Repeat steps (2)-(4). and the damage potential of structures and facilities will certainly change with time resulting in changed stage-damage functions. and describes in detail the basic concepts presented in this chapter. The revised expected annual damage is discounted to the base period and then the raw damage value is amortized over the life of the project to obtain equivalent annual damage. 7-3 . Make certain that the stage datum for the sta~e-dama~e and sta~e-dischar~e functions is consistent for the index location. The differences will be expected annual damage reduction (raw damage reduction benefits) for each plan. at say 10 year intervals with revised evaluation functions at each interval. This adjustment can be of substantial significance. To determine the expected annual benefit it is necessary to account for the changes in expected annual damage that might occur over the life of the project.. To develop a single measure of the damage potential. (3) Combine the discharge-exceedance frequency (in events per year) and discharge-damage function into a damage-exceedance frequency relationship. Equivalent Annual Damaft~. (7) Subtract results of step (4) (with project) for each plan from results of step (4) for without-project measures. i.EM 1110-2-1415 5 Mar 93 (2) Combine the stage-damage and the stage-discharge relations into an intermediate discharge-damage function. (4) compute the area beneath the damage-exceedance frequency relation (expected annual damage) for each index location and sum to obtain the total expected annual flood damage. the expected annual damage must be evaluated over time.

(Note that in practical application there are errors introduced by not knowing the true theoretical distribution of the data. and skew coefficient. Exact statements as to error probability must be based on examination of the frequency curve of errors or the distribution of the errors. (8-1) (8-2) (8-3) These have been used to considerable advantage. thus. Both the standard error of estimate and the confidence limits are discussed in this chapter. Reliability of Freauencv Statistics. standard deviation. it is considered that the standard error is exceeded on the positive side one time out of six estimates. which is defined as the root-mean-square error. One principal advantage of analytical frequency analysis is that there are means for evaluating the reliability of the parameter estimates. they are only approximate for other distributions of errors. and equally frequently on the negative side. for a total of one time in 20. Selected values are given in Table F-9. The standard errors of estimate of the mean. 8-2. the confidence limit curves shown on Figure 8. as discussed in Chapter 9. standard deviation and skew coefficient for regional frequency studies. 8-3.) Criteria for construction of confidence limits are based on the non-central t distribution. In general. An error twice as large as the standard error of estimate is considered to be exceeded one time in 40 in either direction. For instance. which are the principal statistics used in frequency analysis. This permits a more complete understanding of the frequency estimates and provides criteria for decision-making. These statements are based on an assumed normal distribution of the errors. Reliability of Freaue cv Cu es The reliability of analytical frequency determinations can best be~llustr~t~d by establishing confidence limits. for a total of one time in three estimates. are given by the following equations: s~ = S/(N)fi Ss = S/(2N)% SG= (6N(N-1)/[(N-2)( N+l )(N+3)]}% where s~ = S5 = SG= the standard error of estimate for the mean the standard error of estimate for the standard deviation the standard error for estimate for the skew coefficient. in drawing maps of mean. The error of the estimated value at a given frequency based on a sample from a normal distribution is a function of the errors in estimating the mean and standard deviation. Obiectiv~. often termed model error.EM 111O-2-I415 5 Mar 93 CHAPTER8 STATISTICAL RELIABILITY CRITERIA 8-1. Using that appendix. a common statistical index of reliability is the standard error of estimate. and S and N are defined in Section 3-2.I 8-1 .

.. there is one chance in 20 that the true value for any given frequency is greater than that indicated by the ..’” .. nine chances in 10 that the true value lies between the . There are.. ... therefore.’ .95 curve.9 ! Figure 8-1..U’” . & .05 and . Frequency Curve with Confidence Limit Curves.. ~+ . / / / + Qb . .EM 1110-2-1415 5 Mar 93 were calculated. . / ~ ““ .’ .95 curves.. ...... .05 curve and one chance in 20 that it is smaller than the value indicated by the . r 7- . While the expected frequency is that shown by the middle curve. .. 8-2 .’ ~. . ‘*-1968 o — ---- ObeeNed Annual Poeka Computut Frequency Cuwe With Expectd Probability Adjustment Confidence Limit Curvee ..P9’” + ........ 99:9a 99... Appendix E and Example I in Appendix 12 of Bulletin 17B (40) provide additional information and example computations.

EM 1110-2-1415 5 Mar 93 CHAPTER9 REGRESSION ANALYSIS AND APPLICATION TO REGIONAL STUDIES 9-1. Regression analysis is the term applied to the analytical procedure for deriving prediction equations for a variable (dependent) based on given values of one or more other variables (independent). b. the volume of spring-season runoff from a river basin (dependent variable) might be correlated with the depth of snow cover in the watershed (explanatory variable). However. regression analysis will generally permit a more reliable determination of the relation and has the additional advantage of providing a means for evaluating the reliability of the relation or of estimates based on the relation. a. a. Recorded values of such variables over a period of years might be graphed and the apparent relation sketched in by eye. SYand SXare the respective standard deviations and R is computed by 9-1 . The dependent variable is the value sought and is to be related to various explanatory variables which will be known in advance. the linear regression equation is written: Y = a+bX (9-1) in which Y is the dependent variable. The coefficient “b” is evaluated from the tabulated data by use of the following equations (9-2a) or b = RSY/SX (9-2b) in which y is the deviation of a single value yi from the mean (~) of its series. General. and curvilinear regression is therefore not discussed herein. but linear regression suffices for most applications. “a” is the regression constant. and which will be physically related to the dependent variable. x is similarly defined. Nature and At)Dlication. ~alculation of Regression EauationS. The function relating the variables is termed the “regression equation. Regression equations can be linear or curvilinear. and “b” is the regression coefficient. Def inition~.” Correlation is a measure of the association between two or more variables. For example. or explanatory. X is the independent variable. ~imt)le Re~re ssion.” which is the square of the “correlation coefficient. 9-2. In a simple regression (one in which there is only one independent. Often a curvilinear relation can be linearized by using a logarithmic or other transform of one or more of the variables.” and the proportion of the variance of the dependent variable that is explained by the regression equation is termed the “coefficient of determination. variable).

the appropriate set of simultaneous equations can be easily constructed after studying the patterns of the above two sets of equations. the b coefficients can be evaluated from the tabulated data by solution of the following simultaneous equations xx1)2b1 + x(x1x2)b2 + xx1x3)b3 = (Yxl) (9-6) Ix1x2)b1 + Xx2)2b2 + Ex2x3)b3 = (YX2) ~(x1x~)bl + ~x2x3)b2 + ~x3)2b3 = (Yx~) For cases of more than three explanatory variables. ~(yx) and ~(xlxz) can be determined by use of the following equations: 9-2 .. the quantities 1(x)2..-b~~~ (9-7) In Equations 9-2.. . In such cases. Also. and considerable time can be saved by use of the Crout method outlined in reference (51) or (52). the regression coefficients are evaluated from the tabulated data by solution of the following simultaneous equations fixl)zbl + ~(x1x2)b2 = ~(Yxl) xx1x2)b1 + zx2)2b1 = XYX2) (9-5) In the case of three explanatory variables. solution of the equations becomes tedious. programs are available for solution of simple or multiple linear regression problems on practically any type of electronic computer. . Multinle Re~res sion. 9-5 and 9-6.. the regression constant is determined as follows a=~-bl%l-b2~2. b. In a multiple regression (one in which there is more than one explanatory variable) the linear regression equation is written Y=a+blX1+ b2X2 .EM 1110-2-1415 5 Mar 93 Equation 9-11.+ bMXN (9-4) In the case of two explanatory variables. The regression constant is obtained from the tabulated data by use of the following equation: All summations required for a simple linear regression can be obtained using Equations 98 and 9-9a. For multiple regression equations.

with a correlation coefficient of 0.75 (or 87 &rcent) multiplied by the original standard deviation of the dependent variable. The correlation coefficient is the square root of the coefficient of determination.EM 1110-2-1415 5 Mar93 (9-8) (9-9a) X(X1X2) = ~(x1x2)- z xl ~ X2/N (9-9b) 9-3. c.0.5. which is the proportion of the variance of the dependent variable that is explained by the regression equation. which is the highest theoretically possible and indicates that whenever the values of the explanatory variables are known exactly. Standard Error. On the 9-3 (9-12) . Equation 9-10 resolves to R2 = ~(yX)2/~(y)2 ~X)2 (Rz) R2 = (9-lo) (9-11) An unbiased estimate of the coefficient of determination is recommended for most applications. Thus. The adjusted standard error (S=) of a set of estimates is the root-mean-square error of those estimates corrected for the degrees of freedom. the corresponding value of the dependent variable can be calculated exactly. the average error of estimate would be 87 percent of the average errors of estimate based simply on the mean observed value of the dependent variable without a regression analysis. is obtained by subtracting the number of variables (dependent and explanatory) from the number of events tabulated for each variable. and is computed by the following equation: R2 = l-(1 -R2)(N-1)/df The number of degrees of freedom (df). + bMfiYxn) XY)2 In the case of simple correlation.5 would correspond to a coefficient of determination of 0. Dete rmination Coe fficien~. A correlation coefficient of 1. The remaining variance (error variance) would be 75 percent of the original variance and the remaining standard error would be the square root of 0. m..0 would correspond to a coefficient of determination of 1. A correlation coefficient of 0.. a. b. The Co rrelation Coefficient and Standard Error. The sample coefficient of multiple determination can be computed by use of the following equation: bl ~Yxl) + b2 ~YX2) . which would indicate that 25 percent of the variance is accounted for and 75 percent unaccounted for by the regression equation.25.

(9-13a) (9-13b) (9-13C) (1-F2)S. the first requirement of a 9-4 .. ~(Yxl) .8 based on a simple linear correlation with 12 degrees of freedom could come from a relationship that has a true value as low as 0. therefore the flows for the short record station (Tallulah) are assigned to Y. It shows. therefore. The adjusted standard error or error variance of estimates based on a regression equation is calculated from the data used to derive the equation by use of one of the following equations: ~(Y)2 . ~eliabilitv. There is some chance that any correlation is accidental. -.1) a s: . for example.b. that an unadjusted correlation coefficient (R) of 0. the actual standard error of an estimate based on one or more extreme values of the explanatory variables is somewhat larger than is indicated by the above equations.b2 XYX2). a. but this fact is usually neglected. Accordingly. The adjusted error variance is the square of the adjusted standard error. an unadjusted correlation coefficient of 0.8 or lower. and as low as zero in a seven-variable multiple correlation.. On the other hand.. could come from a relationship that has a true value as low as zero in one case out of 20. simD1e Linear Regression ExamDl~. Also. The values in the table are the annuaJ peak flows for the water years 1965-1985 (21 values). The long record station is the Chattooga. about one out of three estimates will have errors greater than the standard error and about one out of 20 will have errors greater than twice the standard error. extreme care must be exercised in the use of multiple correlation in cases based on small samples. b. so the flows for this station are selected as X. The data for this example are the concurrent flows at two stations in Georgia for which a two-station comparison is desired (see Section 3-7). With only 4 degrees of freedom.97 would one time in 20 correspond to a true value of 0. Inasmuch as there is some degree of error involved in estimating the regression coefficients. the less is the chance that it would occur by accident. Phvsical Relationship.53 in one case out of 20. the reliability of a regression equation decreases as the number of independent variables increases. the same unadjusted correlation coefficient based on a multiple linear correlation with the same number of degrees of freedom but with seven independent variables. 9-4.EM 1110-2-1415 5 Mar 93 average. as indicated by the determination coefficient or the standard error. In addition to considering the amount of variance that is explained by the regression equation. but the higher the correlation and the larger the sample upon which it is based. in the case of simple correlation.bn ~(Yxn) df (1-~2) ~(y)2/(N. Ezekiel (8) gives a set of charts illustrating the reliability of correlation coefficients. These two stations are less than 20 miles apart and are likely to be subject to the same storm events. it is important to consider the reliability of these indications. d. An example of a simple linear regression analysis is illustrated on Figures 9-1 and 9-2.

989449 3.43393)(1.749736 4.267171 4.61486 i= 3.——— Chattooga F1OU Year x’ River Log x Iallulch F(OW 71 River —-----Log Y 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 27200 13400 15400 5620 14700 3480 3290 7440 19600 6400 6340 18500 13000 7850 14800 10900 4120 5000 7910 4810 4740 4.802089 4.871572 3.93995 i= 3. (4.459392 3.898176 3. q quation .79069)(3.= 2_356~.65892)(21-1)/(21-2) (by q quation 9-12) = 0.392696 3.0.871572 4. 287.55075 ZY = 73.37213 + 0.79049x (by equation 9-1) logarithms) y.j3351/I.25371 xx ZY .292256 3.614897 3. ——.47956 (X XY)2 = 288.113943 3.668385 3.15&6 )(1.006321 XX = 82.187520 3.79 (without Figure 9-1.658290 - )2/(1. .64031 Contputationa for standard qrror: )/(21-1) (by equation 9-13b) S 2 = (1-0.43393 Xy = 1.127104 4.682145 3.491361 3.8 9-8) 9.07071 XX2= 325.26008 N X2 = 1.795184 3.13351 Y2 = (by (.301029 3.698970 3.167317 3.13351 0.50602 n ( ZY)2 —-— N .36115) (by q quation 9.9a) 1.36115 Regression q quation: Y = 0.292256 3.894869 4.037426 3.02648 se = 0.11) Ez=l (1-0.447158 3. 9-5 .814913 3.382017 3.929418 3.303196 2.79049 a s 3.517195 3.&33922 (by q quation q quation 9-2a) = 0.26115 9-8) b z j. Z748L (Z X)2 — = 324.204119 3.675778 7460 5140 2800 3100 2470 2070 976 2160 8500 4660 2410 6530 3580 4090 6240 1600 t960 3260 2000 1010 2880 3.93099 ZY2= 255.513217 3.37213 R2 s s (1.434568 4.170261 4.541579 3.553883 3.710963 3. Computation of Simple Linear Regression Coefficients.. 254.611723 3.806179 3..334453 3.47956 - (0.640312 q = 0.EM 1110-2-1415 5 Mar 93 .93099) (by 9-3) = 0.

as illustrated on Figure 9-1. This curve represents the best estimate of what the annual peak Tallulah River would be given the observed annual peak on the Chattooga River. using equations given in Section 9-2. The annual peaks for the each station are plotted against each other on Figure 9-2. Illustration of Simple Regression.EM 1110-2-1415 5 Mar 93 regression analysis (logical physical relationship) is satisfied. Re~ress ion Eauation. Because runoff is a multiplicative factor of precipitation and drainage area. Curve B represents the regression line for estimating the annual peak flow for the Chattooga River given an observed annual peak on the Tallulah River. Although not computed in Figure 9-1. 9-6 . The regression equation is plotted as Curve A on Figure 9-2. c. A linear correlation analysis was made. ~&3 *~4 CHATTO06A RIVER ANNUALPEAK (X’ curve A Curve B Regression 1tne with ~ ss dependent 1 ine uith X ss dependent vsriabke verieble Regression Figure 9-2. the logarithmic transformation is likely to be appropriate when comparing two stations with different drainage areas.

156. In this case. If appreciable errors exist in the values of an explanatory variable. If the effect of measurement errors is appreciable. then the regression equation would be more reliable than is indicated by the standard error of estimate computed from Equation 9-13. Consequently. Reliability.3c) and to adjust the correlation results from such effects. Where there is large memurement error of the dependent variable. and erroneous estimates will result. a change in the regression line is effected. If an appreciable portion of the variance of Y (dependent variable) is attributable to measurement errors and these errors are random. Other Factorq. the regression coefficient and constant will be affected. -. If well over half of the variance of the points from the best-fit line is attributable to measurement error in the dependent variable. There is a 95 percent chance that the true value of the dependent variable (Y) for a single observed independent value (X) will lie between these limits. Because logarithms are used in the regression analysis.0 (perfect correlation) consist of pertinent factors not considered in the analysis and of errors in the measurement of those factors considered. and the antilogarithm of twice this quantity is 2. and vice versa. the station having the highest short-time intensity would be expected to have something less than the highest value of mean annual precipitation. the curve is generally closer to the true values than to the erroneous observed values. Thus. it is important that values of the explanatory variables be accurately determined. it is possible in some cases to evaluate the standard error of measurement of each variable (see Paragraph 9. In such a case. if possible. a. an approximate confidence interval can be established at a distance of plus and minus 2 standard errors from Curve A. some locations with extremely high mean annual precipitation may have maximum short-time intensities that are not correspondingly high. Factors ResDonsible fo r Nondete rmination. Hence.EM 1110-2-1415 5 Mar 93 d. Factors responsible for correlations being less than 1. c. Measurem ent Errors. In addition to the curve of best fit.05. if mean annual precipitation were made the dependent variable.05. the effect of adding (or subtracting) twice the standard error to the estimate is equivalent to multiplying (or dividing) the annual peak values by the antilogarithm of twice the standard error. 9-5. by interchanging the variables. As there is a considerable difference in the two regression curves. but would in general have something less than this. In the example used in Section 9-4 there may well be factors responsible for brief periods of high intensities that do not contribute appreciably to annual precipitation. the station having the highest mean annual precipitation would not automatically have the highest short-time intensity. it is important to use the variable whose value is to be calculated from the regression equation as the dependent variable in those cases where important factors have not been considered in the analysis. values of annual peak flow represented by the confidence interval curves are those of Curve A multiplied and divided respectively by 2. the standard error is 0. the standard error of estimate should be obtained by taking the square root of the difference of the error variance obtained from Equation 9-13 and the measurement error variance. The confidence interval is not correct for repeated predictions using the same sample (6). Hence. 9-7 . Therefore. b. On the other hand. Curve B of Figure 9-2 is the regression curve obtained by interchanging the variables Y and X. then the regression line would actually yield a better estimate of a value than the original measurement. This is because the departure of some of the points from the regression line on Figure 9-2 is artificially increased by measurement errors and therefore exaggerates the unreliability of the regression function.

Partial Correlation. In this case. that variable should be used as the dependent variable. and the equation is consequently fairly reliable. the volume of spring runoff is correlated with the water equivalent of the snow cover measured on April 1. however. Its errors will then not affect the slope of the regression line. a. is not ordinarily desirable. The loss in correlation by omitting that variable is expressed in terms of the partial correlation coefficient. In cases where all pertinent variables are considered and most of the measurement error is in one variable. and the resulting difference in slope of the regression lines is entirely artificial. The computations in Figure 9-3 were made with the HEC computer program MLRP (reference 50). Although loss of 4 degrees of freedom of 12 available. Here again. as in this case. b.EM 1110-2-1415 5 Mar 93 d. Reliability of the results. an average slope should be used. Avera~e SloD~. the standard error is about 9 percent. The value gained by using any single variable (such as April precipitation) in a regression equation can be measured by making a second correlation study using all of the variables of the regression equation except that one. H. As discussed in Paragraph 9-3d. questions such as the following should be considered “Would an increase in snow cover contribute a greater increment to runoff under conditions of high ground water (wet ground conditions) than under conditions of low ground water?” If the answer is yes. If it is obvious that all of the pertinent variables are included in the analysis. the standard error of an estimate as shown on Figure 9-3 is approximately 0. An example of a multiple linear regression analysis is illustrated on Figure 9-3. which. c. 9-7. Logarithms should be used for the explanatory variables when they would increase the linearity of the relationship. and the 1-in-20 error is roughly 18 percent. Fun ction of Multit)le Regress ion. is equivalent to multiplying that value by 1. ~. Thus. it was determined that logarithms of the values would be used in the regression equation. the winter low-water flow (index of ground water) and the precipitation falling on the area during April. then a logarithmic dependent variable (by which the effects are multiplied together) would be superior to an arithmetic dependent variable (by which the effects are added together). An average slope can be obtained by use of the following equation b = SY/SX 9-6. Lo~arithmic Transformat ion. In this case. The square of the partial correlation coefficient is obtained as follows: (9-14) 9-8 . In determining whether logarithms should be used for the dependent variable as above. the calculated correlation coefficient may be accidentally high.94) is particularly high. then the variance of the points about the regression line is due entirely to measurement errors. It should be recognized that multiple regression performs a function that is difficult to perform graphically. when added to a logarithm of a value. the adjusted correlation coefficient attained (0. is highly dependent on the availability of a large sampling of all important factors that influence the dependent variable. In other cases where all pertinent variables are considered.038.09. however. Usually logarithms should be taken of values that have a natural lower limit of zero and a natural upper limit that is large compared to the values used in the study.

0000 1. 2W STATISTICS OF OATA AVERAGE .385 .081 .316 .223698 R SWARE .321 .892 1.707 1.460 .0459 .4005 .6308 LOG GW .0572 .127s .2011 .2960 .0531 .9226 STANDARD ERRU OF ESTIMATE .325 .021 .280 .939 LOG SNO LOG GkJ .428 .091 1.2392 .755 .710 .012912 .4170 LOG PRCP LOG Q VARIABLE LOG SNO LOG W LOG PRCP LOG Q -.0000 .634 .0000 REGRESSION RESULTS PARTIAL INOEPENOENT VARIABLE LOG SNO LOG GU LOO PRCP REGRESSION COEFFICIENT 1.060 .6816 .0459 . Example Multiple Linear Regression Analysis.354 . 9-9 .097 .168 .511 .0000 -.0375 Figure 9-3.343 .581 1.0704 .0181 STANDARD OEVIATIDN .9106 .0000 .413 .315 1.744 .45 1946 1947 LOG Q .379 .6308 .0028 .920 .376 .399 .9421 .1346 DEPENOENT VARIABLE VARIABLE LOG SNO LOG W LOG PRCP LOG Q UNBIASED CORRELATION COEFFI Cl ENTS (R2 LOG SNO 1.6170 .2~390 DETERMINATION COEFFICIENT .0000 .266 .621806 1.410 LOG PRCP .127s 1.240 1.295 .038 .297 .2011 1.369 .960 .979 1 2 3 4 5 6 7 8 9 10 11 12 .395 .886 .7451 UNBIASEO REGRESSION CONSTANT -.408 .9196 VARIANCE .666 1.945 1.027 1.9437 R SWARE .EM 1110-2-1415 5 Mar 93 INPUT DATA 00s No OBS ID 1936 1937 1938 7939 1940 1941 1942 7943 1944 19.0050 .181 1.052 .

it must be kept in mind that beta coefficien~ indicate partial correlation only approximately. the other will take over some of its weight in the equation. and when one is removed from the equation. This is done simply by comparing the values of the dependent variable observed. For this reason. = bnSJSY (9-16) The beta coefficients of the variables are proportional to the influence of each variable on the result. These two coefficients are related closely only when there is no interdependence among the various explanatory variables. 9-10. This is particularly helpful for determining whether a relationship is linear and in selecting a transformation for converting curvilinear relationships to linear relationships. and their root-mean-square is an estimate of the standard error of the regression-equation estimates (Paragraph 9-3). but can be found in references 8 and 27. ~ Techniques. particularly when the number of observations is small. This standard error can be compared to that already established in Equation 9-13. For this reason. ~~.(1-Ry:123)/( 1-R. however.12 = 1 . The general theory employed is similar to that discussed above for linear regression. graphical regression methods may prove useful. While the partial correlation coefficient measures the increase in correlation that is obtained by addition of one more explanatory variable to the correlation study. The most important thing to remember in making correlation studies is that accidental correlations occur frequently. V rifi i Acquisition of basic data after a regression analysis has been completed will provide an opportunity for making a check of the results. the regression equation and standard error should be recalculated using the additional data acquired. requires a relatively large number of observations and tedious computations. Practical G uideline$. The differences are the errors of estimate. 9-9. However. beta coefficients are very easy to obtain by use of the following equation: P.EM 1110-2-1415 5 Mar 93 rY:. the beta coefficient is a measure of the proportional influence of a given explanatory variable on the dependent variable. but if the difference is large. variables should be correlated only when there is reason to believe that there is a physical relationship. It should also be remembered that the chance of accidentally high 9-1o . with corresponding values calculated from the regression equation. and the subscripts on the right of the decimal indicate the independent variables. It is helpful to make preliminary examination of relationships between two or more variables by graphical plotting.12) (9-15) in which the subscript to the left of the decimal indicates the variable whose partial correlation coefficient is being computed. 9-8. some explanatory variables naturally correlate with each other. An approximation of the partial correlation can sometimes be made by use of beta coefficients. A satisfactory graphical analysis. Where the relationships among variables used in a regression analysis are expected to be curvilinear and a simple transformation cannot be employed to make these relationships linear. After the regression equation has been calculated. there is no reason to suspect the regression equation of being invalid. Methods used wiil not be discussed herein. If the difference is not signi~lcant.

~raina~e-Basin Charact eristic~. a.Elevation of station . This would be done by correlating important factors with the long-record mean and with the long-record standard deviation of the frequency curve for each station (the long-record values are those based on extension of the records as discussed in Section 3-7). The same principles can be followed using graphical frequency and correlation techniques where these are more appropriate. regional frequency studies may be utilized. Prior to relating these frequency statistics to drainage basin characteristics. Re~ional Freauencv Analvsi$.General slope of surrounding terrain .Elevation of windward barrier . b. In order to improve flood frequency estimates and to obtain estimates for locations where runoff records are not available. if streamflow is correlated with precipitation and drainage area size.Orientation of that slope . If a variable being studied is tested against a dozen other variables at random. Methods of adjusting statistics are discussed in Section 3-7. and then adjust the short-record station values by use of the nearest or most appropriate base station. 9-11. the results of correlation analyses should be examined to assure that the derived relationship is reasonable. -. It might be desirable to adjust the base station statistics by use of the one or two longest-record stations in the region.EM 1110-2-1415 5 Mar 93 correlation increases with the number of correlations tried. Statistics based on precipitation measurements in mountainous terrain might be correlated with the following factors: . For example. even though there may be no physical relation between the two. Freauencv Statistic~. When many stations are involved. because the flow per square mile usually does not increase with drainage area when other factors remain constant. In general. it is best to select long-record base stations for each portion of the region. and the regression equation relates streamflow to some power of the drainage area greater than one. there is a chance that one of these will produce a good correlation. A regional frequency correlation study is based on the two principal frequency statistics: the mean and standard deviation of annual maximum flow logarithms. a maximum exponent value of one should be used. it is essential that the best possible estimate of each frequency statistic be made. A regional analysis involves the determination of the main factors responsible for differences in precipitation or runoff regimes between different locations. Procedures described herein consist of correlating the mean and standard deviation of annual maximum flow values with pertinent drainage basin characteristics by use of multiple linear regression procedures. c.Distance of leeward controlling ridge 9-11 . This is done by adjusting short-record values by the use of longer records at nearby locations.Exposure of gage .

regardless of what fixed values the other independent variables have. as in the case of urban or airport drainage. regardless of the size of drainage area. and in such cases. D. This is not reasonable. If the runoff coefficient is sensibly constant. and runoff. In order to obtain satisfactory results using multiple linear all variables must be expressed so that the relation between the independent and any dependent variable can be expected to be linear. Linear RelationshiD$. regression techniques. If the relation used for correlation is as follows: Q = aD+bA+c (9-17) then it can be seen that one inch change in precipitation would add the same amount of flow.EM 1110-2-1415 5 Mar 93 Statistics based on runoff measurements might be correlated with the following factors: Drainage area (contributing) Stream length Slope of drainage area or of main channel Surface storage (lakes and swamps) Mean annual rainfall Number of rainy days per year Infiltration characteristics Urbanized Area d. There is no simple rule for deciding when to use 9-12 . An illustration of the first condition is the relation between rainfall and runoff. and so that the interaction between two independent variables is reasonable. However. but again a transformation to logarithms would yield a reasonable relation log Q=dlog or transformed D+elog A+logf (9-18) Q= fDdAe (9-19) Thus. if logarithms of certain variables are used. doubling one independent quantity will multiply the dependent variable by a fixed ratio. drainage area. An illustration of the second condition is the relation between rainfall. A. This particular relationship is reasonable and can be easily visualized after a little study. in many cases initial losses and infiltration losses cause a marked curvature in the relationship. Ordinarily. linear correlation of logarithms would be most suitable. Q. it will be found that the logarithm of runoff is very nearly a linear function of rainfall. then runoff can be expected to bear a linear relation to rainfall. regardless of loss rates.

For this reason. the dependent variable is primarily related to the drainage area size. FxamDle of Re~ional Co rrelation. Select ion of Use ful Variables. Equations 8-1 and 8-2 can be used to compute the standard errors of estimating means and standard deviations. correlation analysis for the mean log of annual flood peaks (Y) with several basin characteristics is shown on Figure 9-4. which varies from station to station. but precipitation and slope added a small amount to the adjusted determination coefficient.586 + 0.056 (twice the standard error). In the regression equations shown on Figure 9-4. Many hydrologic variables cannot be expressed numerically. The regression equation selected for the regional analysis included only drainage area as an independent variable. The adopted equation is log Y = 1.EM 1110-2-1415 5 Mar 93 logarithmic transformation. In this example. This increase is because there is a significant increase in the degrees of freedom as each variable is deleted for this small sample of 20 observations. It is usually appropriate. consideration of the increased unreliability of R as discussed in Section 9-3 might indicate that the factor should be eliminated in cases of small samples. there is some loss in determination in only using drainage area. Examgples are soil characteristics. however. 9-13 . vegetation. f. Figure 9-6 shows a map of the errors and Figure 9-5 shows the regional map values for each station and evaluates the worth of the map. but this simple equation is adopted to illustrate regional analysis.0356 for the regression equation alone. In this example.029 or about one chance in twenty that the mean is in error by more than 0. numerical regional analysis will explain only a portion of the regional variation of runoff frequencies. In Figure 9-5 for example. In smoothing lines on such a map. the adjusted determination coefficient increases as variables are deleted according to their lack of ability to contribute to the determination. Both the adjusted determination coefficient and standard error of estimate should be reviewed to determine how many variables are included in the adopted regression equation.839. Use of M~. The simplest equation that provides an adequate predictive capability should be selected.962 log (AREA) or Y = 38. and lines of equal values drawn (perhaps using soils.028. These regression errors are computed by subtracting the predicted values from the observed values for each station. and geology. Combining this regional error with the regression equation should be much better than using the single constant for the entire region. An illustrative example of a regional e. The transformation should provide for near-uniform variance throughout the range of data. Station 5340 (observation 11) had 66 years of record and the standard error for the mean was 0. or topographic maps as a guide). consideration should be given to the reliability of computed statistics.5 AREA”9Q (9-21) (9-20) The ~2 for this equation is 0. The remaining unexplained variance is contained in the regression errors. vegetal cover. when the variable has a fixed lower limit of zero. These errors can then be plotted on a regional map at the centroid of each station’s area. There is about one chance in three that the mean is in error by more than 0. Even in the case of a slight increase in correlation obtained by adding a variable. The map has a mean square error of 0.0112 compared to that of 0.

s00 1.4 115..n44 . . .8 SLWE 6.3 291.0000 . ..1 18.164 -0.IW 1.Oooo .s469 0.230 1.044 4.447 1.0 3...5 3.2053 -.3 MEAN 3..1s6s 0.2m .2 66.2 3.261 1.0540 1.319 1...123 ..0 52.0 1.2655 STAASOARD DEVIATIW .0353 . .0 23.6 34.0 3.0 3.s00 FOREST PRECIP 40.. .4S21 -:% 1...9 3..O 35. .348 3.0 40. -0..5 36.2s12 .l.1s6 3..19ss 0....2749 .370 1..0 1.2s12 := ELEV ..1 24.m 1.025s 0.... ..o 136. -----0.3 22.6 3s.169 -0.6 65..3542 .0 36. .0 65.3635 . .o 2s2. Regression Analysis for Regional Frequency Computations.8 17..0 10.8 39.91s9 LAKES .3 3...9304 .0 81.Oooo ..Oooo H .8 n.269 1.0 2s.ooOo .W 0.6 20..49SS OEPE~NT VARIASLE VMIMLE AREA SLOPE LEUGTPI STW ELEV FOREST PRECIP SOILS m ARRA LEWTH 1.7 U.0 1..130 1.3 44..5s45 .. ..Oooo .W PRECIP SOILS PIEAN 1..6304 .o SO.0 1.0 3.Oooo FORSST ..O 64.s263 .3s32 .1s67 . 0..1187 .0 43.0356 Figure 9-4...000o PRECIP .3s0 .W -1.5 17..0269 0.633 -1 .0 41..5297 .2345 :% -:Z .0020 .%12 1.1096 .2 6. .-..740 1.2WI 1.1939 0.1s9s 0.05s1 .0350 .300 1.0 37.0 S9.0 2:: S3.179 0.o 67..250 0.0 3s..1 3.3W 1.4 22. ..6 1.0000 .2249 . 1s67 ..m 1... .m . -1. .3 ::: U....s5s4 0. .930 3..251 .ss6 0.s 4.2 3..0 298...EM 1I1O-2-I415 5 Mar 93 JIIWT as OATA m as 10 5090 AREA 292.134 1.0 63.07S 3.600 1.. r 43.Oooo .0 1.4304 .2 56..s 1..130s .0302 0..249 OF 0AT4 AVERAOE 2.3635 ..Owo .Owo .m -0.s74 2.Owo .2 3.6327 .%12 SOILS .32s -0....60s -1..s263 ..0000 yRY OF REORESSl~ ASALYSES FOR HEAM LOO OF A-L PEAKS MJUSTEO DETERMISATI~ =FFICIENT sTmMo ERRm OF ESTIMATE MEAs SDUARE ERROR REOSE3Sl~ msTAuT . .2749 -.0 57.2659 . .034 -1.Oooo .8 274..6s77 1.3337 0. .2070 0. 0. .966 t .Oooo 1.7 173.130 -1.0 S2.0 1..1990 0.167 -0.7 21.0 1.267 1.o 114. . . .s097 0.024 1.122 0.301 -0.12ss .023 I.. ...4 3s .0 3s3..169 0.. ..0 34..0 33.7s3 3. AREA EL= LESOTH US ELEV FOREST PRECIP S01 LS (LOO) CLOG) (LOO) (LOD> (ME) (LW) [LOO) tLm) 1.2345 . 9-14 ..0 1..637 3..243 0.129 .2345 l.. . .an ..9 27.4521 .7293 VARIASLE AREA SLWE LENGTH STORAOE ELEV FOREST vARIAscE . .0 37.3 3..3 14.0 443.0707 .0 15 16 17 18 19 20 STATISTICS 5390 5445 54S5 5495 5500 5520 5s25 SOILS 3.0 ELEV 1.0 33.m 1..0 1.364 -0. .8 54.6 37.2 3.Owo .. .030 4.590 4.. .W2 3.0 1.5105 1.0 30.0000 -.350 1.816 4.Oooo ..8 29.8 4s.0 43.2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 51bo 51s0 S200 5205 5260 S270 52s0 S30S S320 S340 53n 53s0 31.Oooo ...000o .5 30.Oow . .7 24.19s .0 20.0 3s.9159 -..2 24.0 85.m .5 215.0 1.s 3..0 604. .0 Trl..6S24 .1S96 0.5 42.0 Is.s297 .4743 .s553 0.140 ----.3s3 1.w 0.. -0..0263 0.Oooo ...5 3..0 1.0 47.027s 0.0 1.200 1..s22 -1.1187 .2S4 3.272 -0.7s3 4..0171 .Oooo .7 LEtiCTN LACES 1.20s3 .165 -0.m .s254 0.0 33.02S7 0.1318 . . .333 3.5230 3.5 2.0 37.054 .4339 1.9 12.122 2....2162 0.1 32.97s O.0257 0.0 3.0 3.Oooo .0 27.7 S. .s390 0.0 la5.700 1.3 1.0 1. REORESS1OS COEFFICIENT .722 S..5 36.0034 ..751 2. law . ..4s9 1.2 17. . . .s366 0.s 1.03s 0.305 1. .Oooo 1.s 15.0 30.25% .m 1..27s 3..0 44.Moo .99s 4.0 S2.104 1.Owo .2 39.0 54.6 22.Woo -?-=7 1.350 1.

035 0.722 3.002 0.590 4.930 3.166 0.029 0.037 0.029 0.930 3.829 4.036 -0.269 3.031 0.256 0.22 -0.019 0.289 0.001024 0.182 0.206 3.07 0.00 -0.020164 0.000676 0.242 0.038 0.02 -0.751 2.333 3.783 4.736 3.028 0.206 0.957 3.275 3.102 3.017 0.08 w DIFF 0.08 0.007 -0.003 D+ 0.000036 0.288 0.W 2.139 -0.078 0.000 -0. Figure 9-6.070 2.047 sun Avor~o Figure 9-5.381 0.17 0.000225 0.766 3.078 3.186 3.o112 39 27 50 35 31 45 44 66 66 40 60 0.05 -0. lb4 0.20 0. Regional Analysis Computations for Mapping Errors.000036 0.145161 0.000081 0. O&b 4.168 0.000004 0.000 0.18 0.000004 0.035 0.278 0.783 3.079 -0.049 0.000225 0.190 0.965 4.005929 0.051 0.142 -0.638 4.109 0.006 0. 9-15 .11 -0.032 0.348 3.15 -0.637 3.006 0.041 0.16 0.341 3.323 0.242 0.000000 0.048 -0.026 0.131 2.174 0. Regional Map of Regression Errors.000036 0.027 0.261 0.738 4.17 -0.601 -0.08 0.015 -0.058 -0.08 0. oo77b4 0.403 3.910 ‘0.050 0.01 0.029 0.006 0.030 4.052 -0.155 -0.224296 0.362 3.055 0.077 -0.195 5260 5270 52B0 5305 5320 5340 5375 5360 5390 5445 5485 5495 5500 5525 0.011881 43 52 0.237 0.092 3.000049 0.09 0.942 3.260 3.186 0.226 41 39 61 39 66 54 39 0.227 0.816 4.102 -0.003 0.068 0.128 0.04 0.339 -0.04 0.251 0.EM 1110-2-1415 5 Mar 93 VAUJ2 1 2 3 4 5 6 7 0 9 10 11 K 13 14 15 16 17 18 19 20 5090 5140 5180 3.187 -0.088 0.002 -0.001296 0.0b6 0.06 0.014 -0.010404 0.039 0.238 3.028 0.291 % 0.164 3.293 0.995 4.015 0.277 0.019321 0.564 3.086 0.029 0.045 0.009 -0.

16). and compute the corresponding determination coefficients. e er Ii Skew coefficients for use in hydrologic studies should be based on regional studies. This procedure eliminates the need to keep the regression constant in the regression equation as the mapped value now includes the regression constant. or adjusted.EM 1110-2-1415 5 Mar 93 h. ~~. after 1950. (10) A frequency curve can be computed for any ungaged basin in the area covered within the mapped re~on by using the adopted regression equations and appropriate map values to obtain X and S. G (4) Calculate z and S for each other station and for the corresponding values of the base station. and draw isopleths of the regression errors for the regression equations of X (see Figures 9-5 and 9-6 fo~ an example) and S considering the standard error for each computed. and then using the procedures discussed in Section 3-2 to compute several points to define the frequency curve. A regional analysis of precipitation or flood-flow frequencies is generally accomplished by performing the following steps: ( 1) Select long-record base stations within the region as required for extension of records at each of the short-record stations. (9) Compute the regression errors for each station. Note that an alternate procedure is to add the regression constant to each error value and develop a map of this combined value. Figure 9-7 is a plot of skew coefficients sequentially recomputed after adding the annual peak for the given year.) i. and tabulate the values for each drainage basin or representative area. (2) Tabulate the maximum events of each station. the skew coefficient was at a at a minimum of about 0. if appropriate. (It may also be necessary to develop regional (generalized) values of the skew coefficient if the Pearson type III distribution is considered appropriate. or one with fewer variables if the adjusted determination coef flcient is nearly the same. plot on a suit~ble map. (5) Adjust all values of% and S by use of the base station. and calculate the correlation coefficient (Equation 3.) (6) S~ect meteorological and drainage basin parameters that are expected to correlate with X and S. (8) Eliminate variables in turn that contribute the least to the determination coefficient.5 in 1954 and maximum of about 1. (Equations 3-1. Su mmarv o fProcedur~. (7) Calculate the regression equations relating ~ and S to the basin characteristics. and select the regression equation having the highest adjusted determination coefficient. recomputing the determination coefficient each time.9 in 1955. only one year 9-16 . using procedures explained in Section 9-2. the longer-record statistics should be used for all subsequent adjustments. 3-2 and 3-3) for each base station. (3) Transform the data to logarithms and calculate ~. Note that. Values based on individual records are highly unreliable. S and. X and S. (If any base station is first adjusted by use of a longer-record base station. The next section describes the necessary steps to compute a generalized skew coefficient. (Equations 3-17 and 319).

(4) a prediction equation should be developed to relate the computed skew coefficients to watershed and climate variables. urbanization. The procedures for developing generalized skew values are generally set forth in Bulletin 17B (pages 10.15). or at least all stations surrounding the area within 100 miles should be included. 9-17 . if possible. (3) the skew values should be plotted at the centroid of the basins to determine if any geographic or topographic trends are present. care should be taken to select stations without significant man-made changes such as reservoirs. (2) at least 40 stations be used in the analysis. and (6) then select the method that provides the most accurate estimation of the skew coefficient (smallest mean-square error). (5) the arithmetic mean of at least 20 stations.EM 1110-2-1415 5 Mar 93 apart. it is recommended that ( 1) the stations used in the study have 25 or more years of data. etc. In addition to the above guidelines. ln summary. in an area of reasonably homogeneous hydrology should be computed.

— — — .— — — — — ul — d m w — 1 I 00 — - - - - — - m I — .

floods originating in a mountainous area or the northern part of the United States at a given site could be caused by melting snow or by rain storms. Procedure. n curves may be more easily combined by the following form: 10-1 . causative conditions.” Also. a. 10-2. In these situations. the equation is Pc = P. A frequency curve representing the events caused by one of the climatic conditions may have a significantly different slope (standard deviation) than for the other condition.EM 1110-2-1415 5 Mar 93 CHAPTER 10 ANALYSIS OF MIXED POPULATIONS 10-1.” For two curves. Bulletin 17B states. but independent. The frequency relations for each separate population can be derived by the graphical or analytical techniques described in Chapter 2-and then combined to yield the mixed population frequency curve. a frequency curve derived by combining the frequency curves of each population can result in a computed frequency relation more representative of the observed events. + P2 . = = c. Annua] exceedance proba bilit v of same magnitude selected above for population series 2. For more than two population series. ~efinition. The individual annual frequency curves are combined by “probability of union. “If the flood events that are believed to comprise two or more populations cannot be identified and separated by an objective and hydrologically meaningful criterion. The largest annual event is selected for each causative condition. A frequency plot of the annual events. PA. floods can be caused by general cyclonic storms or by intense tropical storms. For example.” b. irrespective of cause. Along the Gulf and Atlantic coasts. “Separation by calendar periods in lieu of separation by events is not considered hydrologically reasonable unless the events in the separated periods are clearly caused by different hydrometeorologic conditions. As Bulletin 17B (46) cautions. Figure 10-1 illustrates a combined annual-event frequency curve derived by combining a hurricane event frequency curve with a nonhurricane event curve for the Susquehanna River at Harrisburg. the record shaIl be treated as coming from one population.PIJ’2 where: Pc = PI P2 (10-1) Annual exceedance probab ilit v of combined populations for a selected magnitude. in a hydrologic context. The term mixed population. Annual exceedance probability of same selected magnitude for population series 1. may show a rather sudden change in slope and the computed skew coefficient may be comparatively high. is applied to data that results from two or more different.

For example. Derivation of generalized skew relations for each series can involve much effort. . the frequency curves of two or more tributary stations cannot be combined by the above equation to derive the frequency curve of a downstream site. ~aution$.. (P”)”) (10-2) NOTE: Theexceedance probability (percent chance exceedance divided by 100) must be used in the above equations. This assumes that the events in both series are hydrologically independent. Extensive regionalization may be necessary to reduce the probable error in the frequency relations which results from small sample sizes. d. Coincidental frequency analysis techniques must be used where dependence is a factor. If annual flood peaks have been separated by causative factors. A basic assumption of this procedure is that each series is independent of the other.e. Plate 1 of Bulletin 17B or any other generalized skew map based on the maximum annual event. The lower end of the curve will have a partial duration shape as many small events have been included in the analysis. irrespective of cause. Some series may not have an event each year. the equation is simply PC = PI + Pz. tropical storms do not occur every year over most drainage areas in the United States.EM 111O-2-I415 5 Mar 93 Pc = 1 -(i-Pi) (l-P2). 10-2 . the events in both series must also be economically independent. The combined curve will very likely flt the annual curve only in the middle parts of the curve. quarterly or monthly. Also. and quite often there are only a few flood events for the series. Sometimes frequency relations of particular seasons are of interest. a generalized skew must be derived for each separate series to apply the log-Pearson Type 111distribution as recommended by Bulletin 17B. d. When the combined curve is used in an economic analysis. c. and the curves are combined to verify the annual series curve. it is possible that the slope of the frequency relation will be higher at the upper end of the curve as the one season or month with the maximum event included in its series will likely have a higher slope than that of the annual series. a.. will not be applicable to any of the separated series. b. 10-3. If partial duration curves are to be added. i. This is because the downstream flow is a function of the summation of the coincident flows on each of the tributaries. For instance.

1 .Wx u 0 n u — ‘1 9s.Nonhu=iemeSeriem AnnualHurricune Events. AnnuslNonhumcsneEvent#.1929-8S Computed Frequeney Curve. qh Figure 10-1.9 99 se se 1s 1 . Hurricane. . 10-3 .EM 1110-2-1415 5 Mar 93 o — A —.19 Z9-85 Gr*phicdFrequency Curve.9s 9s. and Combined Flood Frequency Curves. Nonhurricane. = 9s9 se Pwmmt se c~E ie x~-= 1.Hurrie-e Seriee f ~ “/ / u 0 / / I 99.

In Region I. main river stage. ~. A Procedure for Co incident Freauencv Analvsi$. (1) Construct a duration curve for variable B. it is necessary to consider only those events which occur coincidentally with other events. a. The significance of “influential” will be indicated by means of an example. The number of index values of B required for discretization depends on the range of variation of B and the sensitivity of variable C to B. If the variables are not independent. The variable that has the largest influence on variable C is designated m variable A. c. A and B. such data might not be adequate. In the illustration (Figure 11-1 ) the relationship linking variables A. construct an exceedance-frequency curve of variable A. the less influential variable is designated as variable B. a pump station is usually required to pump water only when interior runoff occurs at a time that the main river stage is above interior pending levels. B and C would be obtained with a set of water surface profile calculations for various combinations of main river stage and tributary discharge. develop a relationship between variable A and the combined result C. 11-1 . In some cases.EM 1110-2-1415 5 Mar 93 CHAPTER II FREQUENCY OF COINCIDENT FLOWS 11-1. but it is possible in these cases where the two types of events are not highly correlated to make indirect use of noncoincident data in order to establish a more reliable frequency curve of coincident events. In many cases of hydrologic design. (3) If variables A and B are independent of each other. Index values should represent approximately equal ranges of magnitude of variable B. Figure 11-1 shows water surface profiles along a tributary near the junction with a main river. Procedu re. will tend to have the dominant influence on tributary stage. b. The area under the resulting discretized duration curve should equal the area under the original duration curve. the number of points selected should adequately define the relationships. Determine an exceedance-frequency Variable C is a function of two variables. Discretize the duration curve with a set of “index” values of B. (2) For each of the index values of variable B. Selection of Dominant Variab le. whereas in Region II. The boundary between Regions I and 11cannot be precisely defined and will vary with exceedance frequency. 11-2. Stage on the tributary (variable C) is a function of main river stage and tributary discharge. In constructing a frequency curve of interior runoff that occurs only at such times. Introduction. For example. tributary discharge will tend to dominate. Therefore. relationship for a variable C. Stage-frequency determinations will be least accurate in the vicinity of the boundary where both variables have a substantial impact on the combined result. data selected for direct use should be limited to that recorded during high river stages.

Figure 11-1. Once seasonal exceedance-frequency curves have been obtained (step e). The duration of frequency curves from steps (1) and (3) are assumed to represent stationary processes. $easo nal Effects. d. NORMAL STAGE r@ JBurA. data is generally not available to establish the conditional exceedance frequency curves required by that step. RIVER 9 STAGE NORMAL AUO 100 YR. construct a conditional exceedance-frequency curve of variable C for each index value of variable B. \ 100 VS. they may be combined to obtain an all-season exceedance-frequency curve. Consequently. Repeat this step for other selected magnitudes of C until a complete exceedance-frequency curve for variable C is defined. Assu mntion of Indeoendenc~ Although step (3) enables application of the procedure to situations where variables A and B are not independent. (5) For a selected magnitude of variable C. multiply the exceedance-frequencies from each curve developed in step (4) by the corresponding proportions of time represented. it is assumed that probabilities and exceedance frequencies obtained from the curves do not vary with time. nfGIoll II + SEGIOU 1 100 Yt. In order for this assumption to be reasonably valid. This step is an application of the total probability theorem. On TSIS. MAIII RIVER STAGE MAIM RIVER ASO PROFILE 2 100 O 011 TR IS. STA8E 0 on FOa Tsls. SORNAL RIVER NAI II STAGE qROFILE 1 100 YR. MAIM RIVER STAGE a on FOR Tals. AIAIU TR. e. That is.. Illustration of Water Surface Profiles in Coincident Frequency Analysis.EM I11O-2-I415 5 Mar 93 construct a conditional exceedance-frequency value of variable B. curve of variable A for each index (4) Using the relationship developed in step (2) and frequency curve(s) developed in step (3). NAIU TR. it is generally necessary to follow the above procedure on a seasonal basis. RIVER STAGE qROFILE 3 100 AUO NORMAL O ON TR IS. application of the procedure presented here is generally limited to situations where it is reasonable to assume that variables A and B are independent. and sum these products to obtain the exceedance-frequency of variable C. 11-2 .

the benefits and costs associated with each sequence would be divided by 10 and added in order to obtain the “expected” net benefits. Some work on daily streamflow simulation has been done. The design of water resource projects is commonly based on assumed recurrence of past hydrologic events. Modeling of a stochastic process involves the use of the “Monte Carlo” method of adding a random (chance) component to a correlated component in order to construct each new event. Successful simulation of stochastic processes in hydrology has been based generally on the concept of multiple linear regression. b. If a design is tested on 10 sequences of hydrologic events. Hydrologic records are usually shorter than 100 years in length. 12-3. There is often serious question as to whether the extreme event is representative of the period of record. The severity of a long drought can be changed drastically by adding or subtracting 1 year of its duration. In selecting the number and length of hydrologic sequences to be generated.EM 1110-2-1415 5 Mar 93 CHAPTER 12 STOCHASTIC HYDROLOGY 12-1. The more events that are generated. A stochastic process is one in which there is a chance component in each successive event and ordinarily some degree of correlation between successive events. and most of them are shorter than 25 years. b. ADDlications. If the generation is done correctly. The correlated component can be related. Introduction. the hypothetical sequence would have x equal likelihood of occurrence in the future as did the observed record. it is not logical that a design be based on the most extreme generated event. but the results often apply to other hydrologic quantities such as Precipitation and temperatures. By generating a number of hydrologic sequences. It must be recognized that the more hydrologic events that are generated. the less proportional weight each event is given. not only to preceding events of the same series. it is usual] y considered that 10 to 20 sequences would be adequate and that their length should correspond to the period of project amortization. Even in the case of the longest records. Consequently. each of a specified desired length. for example. the most extreme drought or flood event can be far different from the next most extreme event. the stoch~tic process can be simulated. While it is not possible to create information that is not already in the record. but also to concurrent and preceding events of series of related phenomena. a. the more chance there is that an extreme event or combination of events will be exceeded. Work in stochastic hydrology has related primarily to annual and monthly streamflows. but rather on some consideration of the total consequences that would prevail for a given design if all generated events should occur. it is possible to create a much broader base for hydrologic design. In order that some estimate of the likelihood of more severe sequences can be made. a. 12-2. and long sequences of events can be generated. it is possible to use the information more systematically and more effectively. Basic Procedure. where the regression 12-1 . c.

0 L v= 4. Figure 12-1 illustrates the general nature of the process. a low degree of correlation is illustrated. if every estimate of the dependent variable is determined by the regression line (Figure 12-1a). and the standard error of estimate determines the random component.0 0 2. 12-2 .0 1. in order to emphasize important aspects of the process.0 1.0~~’u ‘. It can be seen that.0 1.0 S. In this case. Data Estimation with Addition of Random Errors. it is necessary to add a random component to each estimate (Figure 12-1 b). Data Estimation from Regression Line.0 4. the estimated points would be perfectly correlated with the independent variable and would have a much smaller range of magnitude than the actual observed values of the dependent variable.0 4. 6.0 ~ V* 3.EM 1110-2-1415 5 Mar 93 equation determines the correlated component. 3. 6. In order to avoid such unreasonable results.0- 0.0 0.0 ---a 7 0 0 9 / 0 m 2.0 () 1.1a.0 0.0 0 0 3.o Figure 12-1 b. and this random component should conform to the scatter of the observed data about the regression line.o Figure 12.0 2.0 4.0 0.0 S.

EM I11O-2-14I5 5 Mar 93 12-4. Monthlv Streamflow Model. Therefore. This type of simulation model can be used to generate related monthly streamflow values at one or more stations.b2 = X1. It has been found that the logarithms of streamflows are approximately normally distributed in most cases. For computational efficiency it is convenient to work with deviations from the mean which have been normalized by dividing by the standard deviation. a.xi)/si # (12-2) where: t = Pearson Type 111deviate i = month number j = year number X = logarithm of ~ = flow mean of flow logarithms s = standard deviation of flow logarithms 12-3 .X2 = Z = Sy = standard deviation of dependent variable R = multiple correlation coefficient b. This deviate is sometimes called the Pearson Type 111deviate and can be computed as follows ti = (xi j . mathematical integrity requires that each variable be transformed to a normal distribution. a simulation model for generating values of a variable which can be defined only partially by a deterministic relation is: Y = a + blX1 + b2X2 + ZSY(l-R2)% where Y= a = dependent variable regression constant regression coefficients independent variables random number from normal standard population with zero mean and unit variance (12-1) b1. Multiple linear regression theory is based on the assumed distribution of all variables in accordance with the Gaussian normal distribution. if it is not already normal. In accordance with the above basic procedure.

k (12-4) where: K’ = monthly flow logarithm. they can be further transformed..1 + ~~K~.l. + ~k.k. if to a distribution very close to normal by use of the following—approximate Pearson ‘Type III transform equation: Ki = (6/Gi) {[(Giti/2) where: necessary. If these deviates exhibit a skewness.k = @lK~. month number for value being generated station number for value being generated number of interrelated stations multiple correlation coefficient random number from normal standard population 12-4 ..~ where m is a station not equal to k and b is the regression coefficient. beta coefficient..2 + .EM 1I1O-2-1415 5 Mar 93 c.l + B2Ki’. defined as bl#i#S. + 1]1/3 + 1) + Gi/6 (12-3) K= i= G= t= normal standard deviate month number skew coefficient Pearson Type 111deviate as defined in Equation 12-2 An equation for generating monthly streamflow is: K ~.1K’i. expressed as a normal standard deviate P= i= k= n= R= z.

Note also that one of the independent variables is the flow for the preceding month in order to preserve the inherent serial correlation. this resolves to: K. b. using equations given in Chapter 2. (2) Compute the mean. The random component.1 percent of the mean annual flow. such as 0. starting with the earliest month of generated data. i. it is necessary to add a small increment. set it to zero.. If a value of zero streamflow is possible. (6) Generate standardized variates for each location in turn for each month. the independent variables will consist of concurrent monthly values at preceding stations and preceding monthly values at the current and subsequent stations. Note that Equation 12-5 is very similar to Equation 12-1. i.l~~ # (12-5) d.EM 1110-2-1415 5 Mar 93 For the case of a single station. by use of (5) Arrange the locations in any sequence.l+Zi(l-R/i.R2)%. The flow value in the original units is computed by reversing the transformation process. subtract the mean from each event and divide by the standard deviation (Equation 12-2). 12-5 . e. In each case. A step-by-step procedure for generating monthly streamflows for a number of interrelated locations having simultaneous records is as follows (1) Compute the logarithm of each streamflow quantity. to each monthly quantity before taking the logarithm.e. When this is done. is a random selection from a normal distribution with zero mean and unit standard deviation. ~. from normal standard deviate to Pearson Type 111deviate. (4) Transform these “standardized” quantities to a normal distribution Equation 12-3. (7) Transform each generated value by reversing the transform of Equation 123 with the appropriate skew coefficient. = Ri . multiplying by the standard deviation and adding to mean in order to obtain the logarithm of streamflow. S. become beta coefficients. standard deviation and skew coefficient of the values for each location and each month. and the standard deviation. (8) Find the antilogarithm of the value determined in step (7) and subtract the small increment added in step (1). equals zero. multiplied by the alienation coefficient which is ( 1 . a. If a negative value results. The differences result from using normal standard deviates. does not appear in the random component since it equals 1. This is accomplished by computing a regression value and adding a random component. to logarithm of flow and finally flow value. the regression coefficients.. the regression constant. and compute a regression equation for each location in turn for each month. (3) For each month and location.l K. according to Equation 12-5.

the program has been designed to generate daily flows after the monthly total runoff has been generated by another program. would consist of (1) stochastic generation of 12-6 . and therefore. a. as well as coordinated hydrographs at many locations in a basin. In estimating the missing values. including frequency and correlation characteristics. b. Ordinarily. A model has been developed (51) that combines the coefficients into a few generalized coefficients for the purpose of generating monthly streamflow at ungaged locations.variation between maximum and minimum runoff standard deviation of flows interstation and serial correlations of flows 12-7. 12-5. or 48 values for one station simulation. it is important to preserve all statistical characteristics of the data.EM 1110-2-1415 5 Mar 93 f. the monthly streamflow model requires four frequency and correlation coefficients for each of the 12 months. 12-6. it is capable only of generating flows at a single location and does not provide a totally satisfactory hydrography. Flows for any particular day are correlated with flows for the preceding day and for the second antecedent day. Dailv Streamflow Model. Since it is desired in many reservoir operation studies to use a monthly interval most of the time. ADDliCat ion In Areas of Limited Data. and to perform daily operation computations for only a few critical periods. Generation of daily streamflows can be accomplished in a manner very similar to the generation of monthly streamflow quantities. as indicated in Equation 12-1. A procedure that will give a reasonable shaped hydrography. ) The generalized model considers the following . periods of recorded data at different locations do not cover the same time span. Data Fill In.season of maximum runoff . It is obviously not feasible toaccomplish the above computations without the use of an electronic computer. To preserve these characteristics. For instance. HEC-4 Monthly Streamflow Simulation (51) can be used for this purpose.average runoff . it is necessary to estimate missing values in order to obtain a complete set of data for analysis as described above. it is necessary to estimate each individual value on the basis of multiple correlation with the preceding value at that location and with the concurrent or preceding values in all other locations. (Procedures for determining generalized statistics for use in generating daily flows have not yet been developed. A random component is also required. A computer program. The streamflow generation models discussed so far have assumed that sufficient records were available to derive the appropriate statistics.lag to season of minimum runoff . Although a computer program has been prepared for this purpose.

All mathematical models are simplified representations of the physical phenomena. and (2) usinga precipitation-runoff resulting stream flow. 12-7 . model to derive the 12-8.” however. It is important at this “state of the art. In most applications. While the simulation of stochastic processes can add reliability in hydrologic design. Reliability.EM 1110-2-1415 5 Mar 93 precipitation over the basin. to examine carefully the results of hydrologic simulation to assure that they are reasonable in each case. the techniques have not yet developed to the stage that they are completely dependable. simplifying assumptions do not cause serious discrepancies.

“Storage Analysis for Water Supply. Frederick. ADDlied Re~ ression Analvsi$. Cornell. 40. i 3rd cd. R.. 1966. M. Army Corps of Engineers. Hardison.S. H. L... A. U. “Storage to be Provided in Impounding Reservoirs for Municipal Water Supply”. Washington.. “Rainfall Frequency Atlas of the U. 1959..S.” 10. The Iowa State University Press. Ezekiel. Geological Survey Water-Supply Paper 1543-A.. J. statistical Methods in Hvdrolo~y. 1914. R. 1973. R. 2. p 1539-1559.. Proba bilitv.S. IA. T.. 1. R.. TX. A. N. 1978. Weather Bureau. Crippen. October 1974. 14. J. May 1961. Dixon. Dalrymple. MD. 3rd cd. Geological Survey Open-File Report 78-352. Department of Commerce. “Five-To Sixty-Minute Precipitation Frequency for the Eastern and Central U. R.” Technical Report CRWR-119.S.” U.” U.. D. D. Intr McGraw-Hill Book Co. Draper. 1969.. GPO. U. Techniques of Water-Resources Investigations. Frederick. MD. Hershfield.. REFERENCES AND TEXTBOOKS Beard. Reston. T. “Composite Log-Type 111Frequency-Magnitude Curve of Annual Floods. States. and Karl A. J. Washington. March 1979. Benjamin. VA. 1960. Sacramento. and H. John Wiley & Sons. “Statistical Methods in Hydrology. 7. Geological Survey. al.w Chapter B2 of Hydrologic Analysis and Interpretation. 8.EM 1110-2-1415 5 Mar 93 APPENDIXA SELECTED BIBLIOGRAPHY A.. Methods of Correlation and Repression AnalvsiS. McGraw-Hill Book Co. C. D...1 12.” NOAA Technical Memorandum NWS Hydro. C. 1970. Beard.S. Silver Spring. 11. J. et. R. Jr. 9.. NY. Smith. “F1ood-Frequency Analysis. A-1 . New York. Silver Springs. 3. Hazen. 1977.” Technical Paper No. 6. CA. Washington. R. Massey. Inc.S. 13. L. 5. “Flood Flow Frequency Techniques. 4. Ames. D. John Wiley and Sons.. and Decision for Civil En~ineerS. Fox. H.. GPO.. Trans.. Center for Research in Water Resources.35. Statistics. W.S. June 1977.” U. and C. M.. Austin. H.C. ASCE 77.. and F. “Interduration Precipitation Relations for Storms-Southeast NOAA Technical Report NWS 21. Haan.C. ~. January 1962.C.

. Department of Commerce. U. April 1966.C. Miller. Kite. Techniques of Water-Resources Investigations. .New Mexico.” Monogram SCR-607. 1960. . Geological Survey. “Two.C... McCuen.. Weather Bureau. Beltsville.to Ten-Day Rainfall for Return Periods of 2 to 100 Years in the Hawaiian Islands. F. Geological Survey Open-File Report 77-792..S. 27. James. U. G.. Weather Bureau.. 3rd cd.. 25. Department of Commerce. Miller. 1964. Owen.” Chapture A2 of Hydrologic Analysis and Interpretation. B.ArizonX Volume IX. U.” NOAA Atlas 2. D. available from NTIS.” Methods of Flow Frequency Analysis.S. 52. 20. VA. C. Canada.Washington: Volume X. Washington.. F. . Volume V. T. W. J. 22. Washington.California Silver Spring.. 17. “Two. Springfield. W. Department of Commerce.C.” Technical Paper No. Washington. E.S. et al.Colorado: Volume IV. “Flood Flow Frequency for Ungaged Watersheds .. . .. Van Nostrand Co.S. 1973. A-2 . Inc.. D. 49. and R. 21.S. . B. U.” General Introduction and Hydrologic Definitions. 19. G. C. November 1977. GPO. Iseri. Langbein.C.Idaho Volume VI.C. H. Miller. Schroeder. Mathematics Dictionar y. D. J. 1974.” Bulletin 13. James.” U. Princeton. December 1977.. Washington.C. “Factors for One-Sided Tolerance Limits and for Variables Sampling Plans. and E. Subcommittee on Hydrology. 1968. H. C. Washington.” Technical Paper No. Miller. J. “Flood Frequency and Risk.Wyoming Volume III. “Two. .Oregon Volume XI.. Washington. D..” Application of a Rainfall-Runoff Model in Estimating Flood Peaks for Selected Small Natural Drainage Basins in Texas. GPO. .. B. D.A Literature Evaluation. 23. inter-AgencyCommitteeon Water Resources. 51.. . “Precipitation-Frequency Atlas of the Western U. MD.S.” Inland Waters Directorate. C..to Ten-Day Precipi@tion for Return Periods of 2 to 100 Years in Alaska.. 1965. Sandia Corporation.Montanz Volume 11. Volume 1.. Reston. et al. Weather Bureau.. H. R. “Frequency Curves.. Massey. 1965.S. 1968.S. Department of Agriculture. 26. Geological Survey Water-Supply Paper 1541-A. Techniques of Water-Resources Investigations. and K.” U. D. March 1963. Riggs. Ottawa. 16. Riggs.“Some Statistical Tools in Hydrology. 1968. 24. J. . .EM 111O-2-I415 5 Mar 93 15. F. D. 18.Utah Volume VII.” Technical Paper No. MD.. Washington.” ARS.” Chapter A 1 of Hydrologic Analysis and Interpretations..C. U. GPO. D. F. D. NJ.S. VA 22151.Nevada Volume VIII.to Ten-Day Precipitation for Return Periods of 2 to 100 Years in the Contiguous U.

D. 39.C. U. Washington District. .. U. Paris.EM 1110-2-1415 5 Mar 93 28. “Flood Characteristics of Urban Watersheds in the United States.S.C. D. CA.S. Davis. “Low-Flow Investigations.. Washington. U. CA. Army Corps of Engineers.. CA.C. 1976.S.. CA.S. the Hydrologic Engineering Center. 37. “Double-Mass Curves. Washington. Sacramento District. A. H. Techniques of Water-Resources Investigations. the Hydrologic Engineering Center.152 Technical Report No. V. A-3 29. C. D. Washington. B. Davis. GPO. GPO. “Volume 2 of Hydrologic Methods for Water Resources Development. “Characteristics of Low Flow Volume-Duration-Frequency Statistics.” U.. “Hydrologic Frequency Analysis. A. and C. GPO. June 1960. U. Water-Supply Paper 1541-A. June 1955. 1. France. 33. Searcy J. Army Corps of Engineers.” U. “Flow-Duration Curves. Flood Flow ComDutation.C.” Volume 8 of Hydrologic Engineering Methods for Water Resources Development. September 1959. Washington.” CW.” U. V. Sacramento District. Sacramento. K.. 35. Army Corps of Engineers. 1960. “Reservoir Yield... GPO. Washington. Riggs. D. 1.. April 1975. 32. 31.” Volume 3 of Hydrologic Methods for Water Resources Development.S. July 1958.S. 42.C. Washington. Searcy.. June 1977. 1972. Army Corps of Engineers. Jr.. U.S.” CW-151 Research Note No. Sokolov. April 1972. A. Sacramento. New York District.. W.152 Technical Bulletin No. Davis. 38. 36. The UNESCO Press. Washington. 34.S. Geological Survey. CA.” Chapture B1 of Hydrologic Analysis and Interpretation. “Regional Analysis of Streamflow Characteristics. H. the Hydrologic Engineering Center. D. Sacramento District. V. 3.S Army Corps of Engineers.” CW-151 Research Note No. U.” CW-151 Research Note No. Geological Survey Water-Supply Paper 1541-B.S. 1960. “Stream Flow Volume-Duration-Frequency Studies. “Probability Estimates Based on Small Normal-Distribution Samples. U. “Hydrologic Data Management. Geological Survey. 1973. D. 1983. Thomas.C. Sacramento.S. et al. “Regional Frequency Study. 41. Methods Comt)i]ed from World ExDeriencQ. 1.. Sauer. Upper Delaware and Hudson River Basins. U. Geological Survey. Army Corps of Engineers.” Frequency of New England Floods. Army Corps of Engineers.” CW. Hardison. 30. Riggs. S.C.” Chapter B3 of Hydrologic Analysis and Interpretation. C. March 1959. Washington District. Geological Survey Water Supply Paper 2207. J. CA. U. 2.. D. November 1974..S.” Hydrologic Engineering Center. U.S. Stricker and K. Techniques of Water Resources Investigations. Army Corps of Engineers. GPO. CA. Wilson. O. “Frequency Curves from Incomplete Samples. K. 40. U. Davis. H. Army Corps of Engineers.

June 1977. U. Simulation of Flood Control and Conservation Systems. Army Corps of Engineers. Switzerland. U.” Computer Program 761-X6-L7580.S. Davis. “Expected Annual Flood Damage Computation. Washington. December 1977. 385. Army Corps of Engineers. CA. February.” Bulletin 15. Army Corps of Engineers. CA.S. “Generalized Skew for State of New Jersey :Special Project Memo 480.” January 1987. “Guidelines for Determining Flood F1OWFrequency. September 198i . 49.” ComPuter program 704-G 1-L2020. 46. CA. 47. Army Corps of Engineers. 55. Water Resources Council. CA.m Computer program 723-X6 -L2500. U. Army Corps of Engineers. CA. 56. Geneva.S. “ HEC-5.” Computer Program 723-X6 -L201O.“ Hydrologic Engineering Center. “Guide to Hydrometeorological No.EM 1110-2-1415 5 Mar 93 43. 54. Hydrologic Engineering Center. World Meteorological Organization.C. D. March 1982. Army Corps of Engineers. Davis. CA. Davis. Hydrologic Engineering Center. February 1971.. “Regional Frequency Computation. April 1982 U. WMO No. U.S. Hydrologic Engineering Center. 48. “A Uniform Technique for Determining Flood Flow Frequencies. Monthly Streamflow Simulation. “HEC. COMPUTER PROGRAMS U.S. CA. Hydrologic Engineering Center. Switzerland.S. Army Corps of Engineers.” Computer Program 723-X6 -L2350. Third Edition.” First Edition. “Hydrologic Analysis of Interior Areas. September 1970. CA. “International Glossary of Hydrology. CA. Davis.” WMO B. A-4 . Water Resources Council. Davis. Army Corps of Engineers. D. U. Hydrologic Engineering Center. U.C. “Flood Flow Frequency Analysis. Geneva. Hydrology Committee. Practices.S. 1982. “Hydrologic Analysis of Ungaged Watersheds with HEC. 51. 1974. Davis.S. 44. 1974. World Meteorological Organization. Hydrologic Engineering Center. Davis. Hydrologic Engineering Center.S. Washington. “Computer Program 723-X6. 53.1. Army Corps of Engineers. December 1967. Army Corps of Engineers. July 1972. 45..S.” Bulletin 17B. U. 52. Davis. April 1982. Flood Hydrography Package. U.1. Hydrologic Engineering Center. “Multiple Linear Regression. Hydrology Committee. “HEC-4. 168. 50.L7550.” Computer Program 723-X6 -L2340. Engineering Manual EM 1110-2-1413. Davis.

Runoff Model ‘STORM’.” Computer Program 723-58 -L7520.” under development. U. Overflow. Treatment. “Storage. U. Hydrologic Engineering Center. Army Corps of Engineers. Army Corps of Engineers. “Statistical Analysis of Time Series Data ‘STATS’. CA. Davis. August 1977.S. A-5 . S. Hydrologic Engineering Center. Davis. CA. May 1987. 58.EM 1110-2-1415 5 Mar 93 57.

( 17B) See “Serial Correlation. The most extreme event.EM I11O-2-1415 5 Mar 93 APPENDIXB GLOSSARY These definitions have been collected from four major sources: (l) Guidelines for Determining Flood Flow Frequency ( 17B). ( 17B) The distribution of sample variances drawn from a normal distribution.” Usually refers to the discharge above which independent instantaneous peak flows are collected for a partial duration frequency analysis. (MD) A systematic record which is divided into seDarate continuous segments because of deliberate discontinuation of recording for significant periods of time. reference 46.” (MD) Autocorrelation Base Discharge Biased Broken Record Chi-Square Distribution Class Interval B-1 . (17B) A list of data in order of magnitude. The expected value of a statistic obtained from random sampling is not equal to the parameter or quantity being estimated. in flood. in a low-frequency analysis. reference 48. (2) International Glossary of Hydrology (WMO). A convenient sized interval into which data may be grouped. DEFINITION Analytical Frequency Analysis Annual Event Annual Series Array A predefine method of estimating the parameters that define a selected theoretical frequency distribution. and (4) Mathematics Dictionary (MD). in the year. Used to compute confidence intervals for the population variance estimated from a sample. either maximum or minimum. (17B) A general term for a set of any kind of data in which each item is the maximum or minimum in a year. (3) General Introduction and Hydrologic Definitions (USGS).frequency analysis it is customary to list the largest value first. reference 18. reference 16. the smallest first. The upper and lower bounds of the class interval are called “class limits.

(USGS) Square of the standard error.EM 1110-2-1415 5 Mar 93 Climatic Year A continuous 12-month period during which a complete annual cycle occurs. The difference between the magnitude of an event and the mean of all the events in the sample. Also called “coefficient of skew” or “skew coefficient. See “Percent chance exceedance. depth Function describing the relative frequency with which events of various magnitudes occur.” Duration Curve Error Variance Exceedance Frequency Exceedance Interval . (WMO) Relation of event magnitude to percentage of events exceeding (or not exceeding) that magnitude. arbitrarily selected for the presentation of data relative to hydrologic or meteorologic phenomena.” See “Recurrence interval. expressed as the ratio of the standard deviation to the mean. The climatic year is usually designated by the calendar year during which most of the 12 months occur. Function of the third moment of magnitudes about their mean. (WMO) A period of abnormally dry weather sufficiently prolonged for the lack of precipitation to cause a serious hydrological imbalance.” (USGS) A numerical measure or index of the lack of symmetry in a frequency distribution. (MD) Coefficient of Skewness Coefficient of Variation Confidence Limits Correlation Covariance Cumulative Frequency Curve Deviation Depth-DurationFrequency Distribution Double Mass Curve Drought The first product moment of two variates normalized by their respective mean values. ( 17B) The interdependence between two sets of numbers.” ( 17B) Statistical parameter describing the change of a stochastic variable in time or space. ( 17B) Plot of successive accumulated values of one variable against the contemporaneous accumulated values of another variable. (WMO) A cumulative frequency curve that shows the percent of time that specific values are equalled or exceeded. a measure of asymetry. See “Water Year. Curve showing the frequency relationship of precipitation for a given storm duration. (WMO) Computed values on both sides of an estimate of a parameter that show for a specified probability the range in which the true value of the parameter lies.

EM 1110-2-1415 5 Mar 93 Exceedance Probability Expected Probability Probability that a random event will exceed a specified magnitude in a given time period. A skew coefficient derived by a procedure which integrates values obtained at many locations. ( 17B) The probability of rejecting a hypothesis when it is in fact true. Procedure involved in interpreting a record of events in terms of future probabilities of occurrence. (17B) Level of Significance B-3 . At a “lO-percent” level of significance the probability is 1/10. ( 17B) The Nth root of the product of N values or the antilogarithm the mean logarithm of a set of values. Univariate frequency diagram with rectangles proportional in area to the class frequency.” The number of events in a sample (or the population) that meet specified criteria. (MD) Curve See “Duration Curve. erected on a horizontal axis with width equal to the class interval. usually one year unless otherwise indicated. of F Distribution Flow-Duration Frequency Frequency Analysis Frequency Curve Generalized Skew Geometric Mean Graphical Frequency Analysis The development of a frequency curve by drawing a smooth curve through plotted points while considering known constraints. (WMO) Information about significant events before or after the period of “systematic” data collection. and then plotted on the appropriate probability paper. Plotting positions are computed based on the order number and the total number of values represented. (derived from 17B) Records (samples) from the same population. (WMO) A graphical representation of a frequency distribution. Usually a cumulative frequency curve with the abscissa a probability grid and the ordinate the event magnitude. ( 17B) Histogram Historic Data Homogeneity Incomplete Record A streamflow record in which some peak flows are missing because they were too low or high to record or the gage was out of operation for a short period because of flooding. ( 17B) The average of the true probabilities of all magnitude estimates for any specified flood frequency that might be made from successive samples of a specified size. ( 17B) The random sampling distribution of the ratio of two independent estimates of the variance of a normal distribution.

even though most data are not exactly normally distributed. Not stationary with respect to time.” A probability distribution that is symmetrical about the mean. The most frequent value of a set of numbers. the median is the average of the two middle values. (MD) Mean Daily Median Method of Moments Mixed Populations Mode Non-Central t Distribution A distribution that combines the probable error in the mean and the standard deviation for samples from a normal distribution. ( 17B) Outliers (extreme events) are data points which depart from the trend of the rest of the data. median. (USGS) Nonstationary Normal Distribution Outlier Parameter Parent Population Partial-Duration B-4 . It is the most studied distribution in statistics. Used in the development of confidence limit curves about a frequency curve computed from sample statistics. such as mean or standard deviation. Series A list of flood peaks that exceed a chosen base stage or discharge. Distribution Mass Curve Mean to the logarithms Curve of an accumulative quantity versus time. ( 17B) A sample whose events have come from two or more different populations. Parameter estimates are called statistics. The arithmic mean (or average) of a sample is an estimate of the population mean. See “Stationary Process. See “Population”. i. or the mean of values within one day. (WMO) The expected value of a random variable.EM 1I1O-2-14I5 5 Mar 93 Log-Pearson Type 111 Application of the Pearson Type 111distribution of the data. (derived from WMO) The value at which one-half of ordered observations lie on either side. the Laplacean. It is also known as Gaussian. or the Laplace-Gauss distribution. the first moment. A standard statistical computation for estimating the moment of a distribution from the data of a sample. and mode (bell shaped). because of its value in theoretical work and because many other distributions can be transformed into the normal. If there is no middle value.. or the Second Law of Laplace. the Gauss-Laplace. ( 17B) A characteristic descriptor of the population. The mean of daily values in a specified period. data not homogeneous.e. regardless of the number of peaks occurring in a year.

theoretical frequency distributions which the normal distribution is a special case.EM 1110-2-1415 5 Mar 93 Pearson Type 111 Distribution Percent Chance Exceedance Percent Chance Non-Exceedance Plotting Position Population Family of asymmetrical.) An element. the principle of minimum squared error (least squares) is used in the derivation. the average interval between floods of a given size. or fragment of a “population. that has an assumed “true” percent chance exceedance. Risk is computed by the binomial distribution. ( 17B) Extension of the results of the frequency analysis of point data to an area. The average time interval between actual occurrences of a hydrological event of a given or greater magnitude. during a specified number of years. expressed as a percentage. In a partial duration series. with which values exceed a specified magnitude. and future floods at a location on a river is the population of floods for that location even if the floods are not measured or recorded. In a hydrologic context. of The probability. (For contrast. with which values will not exceed a specified magnitude.” Every hydrologic record is a sample of a much longer record. For example. total number of past. part. The distinction holds even though for large floods. Any event in the population has an equal chance of being selected. of an observed The entire (usually infinite in hydrologic application) number of data from which a sample is taken or collected. see uncertainty. Commonly. recurrence intervals are nearly the same for both series. In an annual flood series. ( 17B) B-5 Probability Random Recurrence Interval Regional Analysis Regression Return Period Risk Sample . expressed as a percentage. ( 17B) The ratio of the number of random events with some particular size or other attribute to the total number of equally likely events. present. the average interval in which a flood of a given size is exceeded as an annual maximum. the probability that one or more events will exceed a given annual event. See “Recurrence Interval. (WMO) An analytical procedure that derives estimation or prediction equations for a variable (dependent) based on given values of one or more other variables (independent).” The probability of a potential outcome (success) being realized within a specified number of events (trials). Percent chance of exceedance (or non-exceedance) value estimated from its position in the array. regardless of their relationship to the year or any other period of time. The probability.

Process in which both the probability and the sequence of occurrence of the variables are taken into account. The most common transformations are those changing ordinary numerical values into their logarithms. It is now standard practice in statistics to divide by the number of values minus one in order to get an unbiased estimate of the variance from the sample data. For all the many types of tests. to linearize a plot or to normalize a skewed distribution by making it more nearly a normal distribution.” A test mode to learn the probability that a result is accidental or that a result differs from another result. See ‘Coefficient of Skewness. there are standard formulae and tables. (WMO) of a statistical Serial Correlation Skew Coefficient Stage Standard Deviation Standard Error Stationary Process Statistic Stochastic Process Student’s tDistribution Systematic Record t-Distribution Test of Significance All of the generating moments of the frequency distribution remain fixed with respect to time. many others are possible. (USGS) A measure of the dispersion or precision of a series of statistical values such as precipitation or stream flow. ( 17B) Standard deviation of the sampling distribution parameter. ( 17B) Information collected by a systematic data collection program. ( 17B) The change of numerical values of data to make later computations easier.EM 1110-2-1415 5 Mar 93 Sampling Error The difference between a random sampling statistic and the parameter of the population from which the random sample was drawn.” The height of a water surface above an established datum plane. An estimate of a population parameter obtained from a sample of the population. Also called autocorrelation. (WMO) A distribution used in evaluation of variables which involve sample standard deviation rather than population standard deviation. square roots or cube roots. It is the square root of the sum of squares of the deviations from the arithmetic mean divided by the number of values or events in the series. (derived from 17B) See “Student’s t-Distribution. ( 17B) B-6 Transformation . (MD) A measure of the interdependence between an observation at a given time period and that of a preceding time period.

EM 1110-2-1415 5 Mar 93 Trend A statistical term referring to the direction or rate of increase or decrease in magnitude of the individual members of a time series of data when random fluctuations of individual members are disregarded. it is the twelve-month period. carryover is reduced to a minimum. use of the term to signify drainage basin or catchment area has come to predominate. Geological Survey reports. (USGS) Continuous twelve-month period selected in such a way that all solid and liquid precipitation runs off during this period. (WMO) In U. A measure of the amount of spread or dispersion of a set of values around their mean.S. October 1 through September 30. and hence equal to the square of the standard deviation. although drainage basin is preferred. Thus. The water year is designated by the calendar year within which most of the twelve-months occur. (USGS) Unbiased Uncertainty Variance Watershed Water Year B-7 . (USGS) The expected value of a statistic obtained from random sampling is equal to the parameter or quantity being estimated. However. obtained by calculating the mean value of the squares of the deviations from the mean. model uncertainty comes from assuming a theoretical frequency distribution and parameter uncertainty comes from estimating the parameters for the selected distribution by sample statistics. In frequency analysis. (MD) The inherent error in an analysis caused by not knowing either the true model or the model parameters. ( 17B) The divide separating one drainage basin from another and in the past has been generally used to convey this meaning. over the years.

6. A commonly used distribution of extreme values (annual series) is the double exponential distribution. In this method Sn (5.1.2 Compdational methods functions applicable to hydrological It can be shown that most frequency analysis can take the form XT. If X is not normally distributed.1) where % is the mean value.2) c-1 .) 5.denotes the magnitude of the event reached or exceeded on an average once in T.2. K is the frequency factor. years.EM 1110-2-1415 5 Mar 93 APPENDIXC COMPUTATION PROCEDURE FOR EXTREME VALUE (GUMBEL) DISTRIBUTION (Reproduced from reference 48. Value XT. = ~ + Ks. K depends on frequency and skewness coe~cient. and often bears his name. (5. which has been widely applied by Gumbel (see Biblio~aphy). and s= is the standard deviation of the variable being studied.

6247 1.1205 3. a~ld sO.8483 1.:522 1.the r~”duccd standard dcviatit)ll.6373 i.6665 1.1393 -0.4023 50 3.4565 2.1470 -0.1422 -0.1496 5 1.[luccd mean.4442 2.9672 .1434 -o.5929 1.8932 3.5754 1.1355 -0.8701 25 2.3208 3. known as extreme probability paper.6993 2. TABLE 5. 10 (5.4).i454 -o. computation of the values of ~ and s= (Table 5 .0743 3.1484 -0.)ne consists in computation of XT.1408 -0. by means of Eq.i510 -o.9265 3.9352 .7766 :.5354 2.8467 2.7894 2.4219 2. is related to return pf.5874 3.8784 .6805 3.9447 .7409 2.1664 1. AN. (5.6512 1.4699 2.EM 1I1O-2-I415 5 Mar 93 liYI)RO1.2270 3.2549 3.71i3 3. and drawing a line by inspection.3 values givf.5676 1.8879 . = . s“.6665 (continued) 20 21 22 23 24 25 26 27 28 29 -0.3227 4. (53) ) }“.9i87 ..5535 4.1040 3.1506 -o.\l.0338 1.8931 .1376 -0.9635 3..9806 .4048 3.5559 2.io50 4.1787 3.9049 .s values of K computed of Eq.5163 3.8988 .6954 3. The other consists in plotting data on suitable ~aph paper.riofl l)> T.83405 ( + 2.6632 2.1501 -0.8356 3. are furlctillns only of salnple size.5169 2.0517 4.3600 3.2379 4.6035 2. L}lI.1478 -0.5603 1.8631 3.i444 4.0610 3.9553 .8106 3.9265 .4843 2.7466 3.2) using Gumbel’s for ~“. There are two basic nlcthssds for fitting data to the extreme value distributio[l.7025 i.OGIC.1383 3.7240 1.2) Return period (years) n 10 11 12 i3 14 i5 16 17 18 19 2 4.r[.5470 c-2 .6026 1.7663 3.5838 1.-) ‘1 1 whcrc y.0886 3.r ‘0.3025!)log log ~ by means Tablo 5.2860 3.4118 2.0368 100 4. 1).1490 -0.6316 2.1576 3.6132 1.5784 2.8742 .0580 1.0485 3.0134 .4999 2.i463 4..8830 .i518 -0.9958 .8094 1. (5.0049 3. at)~l YTr.7875 3.i5i5 -0.3 VSZLUPS o/ K bed on Eq.4326 2.4563 3.6835 1. lhc reduced variate. Y51S . and ~Tr.2017 3.7484 1.*l.after a previous (.7283 3.9115 .

8486 2.1529 -0.4594 1.4352 3.8504 .5467 3.50~6 5.8450 .8228 .349f 2.3211 2.8562 .8807 2.1526 -0.i568 -0.8397 3.8197 .5110 1.3f15 2.6534 36410 3.1545 -0.1575 -0.1571 4.5259 3.5073 3.8182 .4712 3.8937 2.8MB4 .4317 c-3 .8379 .8141 .5325 3.5068 1.4424 3.4458 1.3934 2.4391 25 50 ! 00 30 31 : 34 35 36 : 39 40 41 42 43 4.8476 .4475 1.5881 3.1559 4.1562 -0.4712 1.8583 2.2868 2.4573 1.4824 1.3069 2.8092 .446i 3.4856 3.8168 .1.1557 -0.908~ 2.2699 2.8532 .1564 -0.1572 -o.3026 2.9425 2.8892 2.8080 .2944 2.5028 1.4500 3.1570 -0.9789 2.6076 3.i552 —0.5194 3.ti292 3.1566 -0.8317 .37’70 2.8455 2.8116 .4854 1.5i53 1.4 45 46 47 48 49 50 51 52 53 54 55 : 58 59 60 61 62 63 64 —0.5.2731 2.567 4.2905 2.8357 .8103 .82i2 .859Ii .8849 2.4667 3.1540 -0.4908 3.2763 2.i577 -0.9633 2.3162 2.5353 1.3850 2.8069 .1532 -0.5790 3.2639 2.4663 1.8038 .1550 -0.i556 -0.4639 1.1561 -0.4407 1.8154 .2984 2.8653 2.6425 .4424 1.6181 3.8728 2.1548 -0.85t8 2.1582 -0.8245 .5704 3.5248 i.4954 1.9133 2.1583 --0.4886 1.2529 2.4759 3.9031 2.5299 i.5410 1.2669 2.4387 3.0153 3.EM 1110-2-]415 5 Mar 93 HYDROLOGICAL TABLE ANALYSIS 5.4766 1.1576 -o.2797 2.4990 1.8690 2.4739 1.8618 2.2504 2.4807 3.2432 3.3556 2.5622 3.4794 1.4540 3.8279 .4494 1.8628 .8550 2.2479 2.2556 2.8337 .0257 3.496i 3.1543 -0.2583 2.1583 .4616 1.1581 -0.4920 1.9961 2.2832 2.9873 2.3 (continued) Return period (years) n 0 5 10 1.8262 .1578 -0.0054 2.5976 3.4623 3.337i 2.8402 .43 3.8767 2.9167 2.1579 -0.9709 2.9491 2.5i99 f.8058 .5133 3.3623 2.8028 2.8983 2.3430 2.4687 1.4552 1.2455 2.930~ 2.45i2 i.9561 2.8426 2.8048 .9243 2.3695 2.i573 -0.1554 -0.2610 2.8298 .1535 -0.4581 3.3315 2.5395 3.1538 -0.4532 1.8128 .1580 -0.4440 1.9362 2.3262 2.

1869 .: 77 78 79 80 8: 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 --0.2128 2.i597 -o.7982 .8263 2.1596 -0.4048 1.i590 -0.2026 2.7963 2.3640 3.22fi 2.7840 .4219 3.7851 .1604 1.2344 2.4158 3.8018 .4332 1.7906 .3 (continued) Return period (years) n 0 5 .4207 1.4154 1.4305 1.425i 3.IW2 -o.4oio 2.1596 4.2160 2.3916 3.7862 .4291 1.1937 2.3892 3.cL54 1.7783 2.4145 1.7893 2.4116 1.i588 -0.8167 2.3991 3.1587 4.4284 3.7927 .3823 3.8039 2.3569 3.7899 .8122 2.7819 .1595 -0.4072 1.1603 4.7800 .779i c-4 .7910 2.1599 -o.4089 1.4017 3.7726 2.4128 3.4071 3.i598 -o.8341 2.4 2.1902 2.1595 4.7713 2.1594 -0.i59i 4.19:3 2.7844 2.7950 .3487 65 66 67 68 69 70 71 72 73 74 -.7740 2.i589 -o.1949 2.7829 .3845 3.1585 -0.7965 .1925 2.7834 .22% 2.4056 1.7809 .3678 3.3965 3.1593 4.7893 .7886 .8i44 2.1600 -0.7868 .7880 .7877 2.4107 1.i592 -o.i599 -0.7769 2.4242 1.4185 1.2012 2.i598 4.7913 .8214 2.7934 .7860 2.2365 2.EM II1O-2-1415 5 Mar 93 llYDROLOGICAL TABLE ANALYSIS 5.1604 -0.2409 2.3586 3.2053 2.1592 -0.7982 2.7974 .2265 2.IM 2.7874 .1880 2.8000 2.2143 2.2193 2.8314 2.3535 3.8190 2.4033 1.4044 3.7845 10 25 2.i973 2.7920 .4081 1.7945 2.3758 3.7927 2.i600 -0.8288 2.7754 2.4165 1.4346 1.2097 2.189f 2.3940 3.2082 2.2i76 2.3801 3.1601 -0.7813 2.1587 -o.4218 1.4135 1.4125 1.2067 2.7804 .8000 .2~8.3552 3.8101 2.4025 1.7991 .8368 2.8059 2.2i12 50 2.3503 3.4376 ~.7795 .7828 2.3622 3.4230 1.7957 .1602 -0.3868 3.4318 1.4196 :.7942 .8020 2.1586 -0.2324 2.4i88 3.i593 -0.1597 -o.4098 1.2039 2.4018 i.3659 3.7856 .4361 1.4099 3.23 5.i603 -0.1590 -o.3779 3.7700 100 3.4278 14266 i.3604 3.3717 3.2304 2.3519 3.7824 .1602 -0.7814 .4064 1.4175 1.2246 2.4040 1.8009 .7798 2.3698 3.3738 3.1584 -0.1986 2.2387 2.8238 2. .1601 -0.i96i 2.8080 2.

m being 1 for the largest.3 (5 5) where n is the number of yearn of record (the number of items in the annual series) and m is the rank of the item on the series.24 lIYDROLOCICAL ANALYSIS Extreme probability paper has a linear ordinate for the variable being studied. Plotting positions arecommonly determined bytheformulac[l6]: n+l T.4.9. and the abscissa is a linear scale of the reduced variate (Eq. Rainfall depths for return periods other tharl fO yearn can be computed in a similar manner. the warped scale of T~ is also shown alorlg L}ICtop of Fig.4 m— 0. Computations are illustrated in the lower part of the table for T.4. hypothetical values of a series of annual rainfall maxima are ffven in the upper part of Table 5. c-5 . Tr=n +0. .Example of extreme probability plot using data of Table 5.3)]. . 5. of iO. For convenience in plotting.(5.9 . Toillustrate the stepsin numerical computation of the rainfall value fora @ven return period. .EM 1110-2-1415 5 Mar 93 5. EXTREME PROBABILITY PAPER Figure 5.=— or m (5 4.

and P 10 = 46. and in Fig.900 8. lC.024 3. for T. 1): Pz = 46.2 1’ — F 81 676 196 196 44i 36 0 576 2.8094 X 21. a record of only 11 years is used.7 In Fig.704 2.\L TABLE . K = 1.600 625 2.9 the two + ‘s show the above values for P: and Plo al~d define the line shown. 5.92 SS by short cut: \/ For T.92 = 65. with rainfall values for each plotting position.0 2. In the table.0 — 0.1376 (from Table 5.6094.9.\ NALISIS 5. reference is a-in made to Table 5.0 ~.9 the plotted points are given. 5.0 12. 3).72 i. values of the plotting position are @ven.4 Computation 0/ eztreme values 11+1 Ill m 1.116 4.116 4 4U 4.1376 X 21. for convenience.5 4.0 1. To illustrate the graphical method of fitting data to the extreme-value distribution.0+ i. 5. (5.606 p2 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 37 ~() 32 60 25 52 46 70 92 48 24 7 ii 8 3 9 4 6 ~ i 5 io —9 — 26 — 14 + 14 — 2f +6 o + 24 + 46 +2 — 22 1. = 10. Substituting i (P)*—Fi P = r 4606 — = 21.=2andn=ll.4 and Fig.0 Similarly.EM 1110-2-1415 5 Mar 93 II YDROL{I{. The curve shown could have been drawn by fitting the plotted points by inspection.0 SX by square of deviations: di n —i (P)2 —Fi n— 1 P = F ~ 10 = 21. C-6 .33 3.92 10 K = —0.() 6.269 400 1.082 = 46.4 1.464 2. but for longer return periods the short record has a large sampling error and the computations should not be taken as precise estimates.304 576 Total F=2P/n=~ 506 506 28.09 1.92 = 43. into Eq. Such a record gives a fairly stable value for return periods of as much as five years. In this example.

se ~=r.000 In most cases.f)(.~ (~ can be obtained directly from Table 5.fsdence levels are as follows: a = 95 0/0 =goyo :=80yo a=68~o t(a) t(a) t(a) t(a) = = = = 1. a better fit of the data. However.282 X 0. or a closer approach to linearity. so that a decision is based on differences in data rather than differences in subjective interpretation of data. This advantage has the corolla~ of standard treatment of data. For discussion of additional distributions and of additional methods for fitting distributions.645 1. siderable sampling error. l(:. and of close fit to a particular distribution. one mav exnect to find the true precipitation ‘vah.14K + the following relation [19] exists: l.7 in Table 5. in determining the 80 per cent confidenceinterval for PIO= 85.4. 6) ‘=~’”fi In particular. 18. (5. t(a)% = 1.92 = 21. it should be remembered that extrapolation involves con.960 1. Thus. such as the normal or lognormal Pearson Type III distribution. and periodical statistical literature [i7. values of ~. 63-66].~G }{ YDRO1. the standard by means of the formufa error of estimate. it would be desirable to compute the confidence interval with limits: XT.5. is the facility for extrapolating beyond the range of the data.AI. instead of using the extreme-value distribution. An advantage of fitting data to a distribution is achievement of objectivity.6) values of ~. with eiven confidence levels. A commonly used distribution is one in which the magnitude scale is logarithmic and tile probability or return-period scale is the normal distribution. A third advantage of linearity in plotting points. for the Gumbel distribution ~. This distribution and the plotting paper used with it are widely known as log-normal. 7) where K is the numericalvalue definedby Eq. + t(a)Se within which.10K~ (5.2) and readily obtainable from Table 5. reference may be made to textbooks. — t(a)*.9 c-7 .3.for conveniencein the use of Eq.\L AN. can be computed (5.282 1. values of t(a) for selected” co.EM 1110-2-1415 5 Mar 93 5. may be obtained from one of several other types of distribution. for example. (5. YSI+ For some types of data. However.= \/f + 1.7783X 21. For evaluation of the accuracy of the computed values XT. XT.

5898 .5471 .4006 .5976 (cotiind) .7559 .4815 .9196 .3645 .4314 .5863 .7783 .8134 .7097 . I and 50.6673 .1476 .3121 .5252 1.4702 .5069 .3394 .2811 .6392 .497i 50 1.3857 ./fi /or use in Eq.2681 .1937 i.3698 .6468 .1657 .2018 . the lower and upper limits of the same confidence interval for P* = 43.7358 .7198 .9.1971 .2726 .i631 .8285 .1537 .3007 1. Similarly.4230 .4595 .6257 .8657 .2316 .i927 .3813 .5522 .1777 .5123 .9532 .3648 1.4402 ./\/~ being 0.9064 .9290 .1847 .6983 C-8 .8i36 .3938 .8472 .6142 .9155 .3473 .2275 1.7596 .7750 .3249 .7 + 21.5211 .4496 .2068 .6) Return period (yeurs) n 10 11 12 13 f4 15 16 17 18 i9 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 2 .7692 .1516 .5289 .2803 (Table 5.2479 .0382 1.9 and 85.5 Vdw o/ f3r.5569 .7313 . 5).0i61 .1100 1.EM 1110-2-1415 5 Mar93 HYDROLOGICAL ANALYSIS 5.1582 .4242 .7146 .5232 . TABLE 5.7–21.2393 .9645 .1745 .5532 .7942 .1685 .5918 .4152 . respectively.3874 .0620 1.3873 :.5377 .365i .4389 1.1496 .8370 .7088 .i488 1.6948 .5673 .3559 .3183 .0 arc 35.3062 .4569 .7978 .887i .0716 i.3754 .0075 .6437 . which means that there is an 80 per cent probability that the true value of PIO lies between 63.2803 .4982 .2648 10 25 1.7245 .4937 .7571 .6595 .6i59 .7403 .5365 .2574 .2686 .6.2439 1.2123 5 .5782 .8610 .1811 .5716 .6020 .%56 .3750 .076i i.9467 .i606 .2246 .4102 .6066 .3005 1.8854 .6699 .0193 .6i50 .4397 .1886 .3319 .6825 100 1.6581 .1083 1.5045 .981i .8298 .1714 .27 The lower and upper limits of the confidence interval are therefore 85. respectively.8 and 107.9793 .6289 .6992 .6273 1.3007 . L}lCvalue of ~.3974 .8450 .2i82 .1646 1.4763 .5204 .8148 .7853 .6360 .9.6765 .6957 .2767 .2954 .7361 .7831 .8801 .7433 . (5.1559 .2904 .1472 1.2942 .4077 .

2135 .224i .4487 .5635 .3106 .2953 .7)4~ .2479 .4900 .4228 .4832 . .3611 .5 AS ALYSIS (continued) Ret urn period (years) II 4n ? ~ .5239 .2175 .5426 .2008 .l~go .4690 .1219 .4377 . .5996 .5931 .5373 .324(.~.2115 .4274 .4018 .5580 .2872 .3534 .6064 .5330 .3646 .5231 .1344 .2155 .28 HYDROLOGIC.1179 .1303 .4274 .2218 .4973 .+L TABLE 5.2096 .3365 .~58~ . .1108 .2796 .5527 .2078 .4734 .3905 .2821 .4210 .1960 .3769 .4825 .1277 .2925 .5195 .1092 .~~l .5110 .4584 .2042 . .4914 .1142 .1100 .4990 .4321 .5807 .2364 .1077 .2363 .4183 .2726 .4565 .4873 .3451 .4140 .2982 .ii88 .3507 .3042 .1930 .5808 . .1124 .3706 .1169 .~~4f .1230 .476(.3589 .1116 .4431 .1874 10 .1976 .3441 .3870 .5134 .4525 .~37(.lgo~ .2391 .4643 50 .1070 41 ~~ 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 ~~ 73 74 .3407 25 .3979 .4413 .3324 .5312 .3209 .2772 .4179 .EM I11O-2-I4I5 5 Mar 93 5.5377 .5152 .1992 .1316 .3836 .4606 .5748 .1S8S .4098 .2288 .5580 .3.2449 .4779 .4703 .5284 .2312 .4057 .1405 .3284 .1265 .4952 .4370 .4528 .3012 .2683 .~4i9 .5868 .1150 .5!~3 .3481 .3737 .i084 .2704 .2898 .2749 .26G1 .5025 .2025 .3344 .1330 .1457 .4878 c-9 .191(.1253 .3617 .3139 .5509 .143!~ .5i92 .1242 .138!) .5029 .4449 .1209 .3802 .~G2~ .2~6 .3073 .2510 .4647 .1373 .3676 .5653 .34!)~.1133 .4474 .2060 .6133 .2339 .1198 .1160 .3174 .1422 .4208 .5890 .3942 .42G2 .5690 .1945 .572!) .1358 .5476 .4342 .6205 .2196 .5079 .3561 .2264 .4922 .

3407 .i66i .4142 An examDle of the magnitude of error in extrapolation beyond the range of the data m~y be found ii the record of maximum” annual 24-hour rainfall it Hartford.3569 .2364 .zf~l .A.4118 .3059 90 91 92 93 94 .2425 .1786 w 81 82 83 84 .0969 .0948 .3670 .\ SALYSIS TADLE 5.0934 .2975 .0964 .1810 .1016 .0939 .4187 .4s42 .3337 .1752 .2267 .4164 .3649 . In 1955 a hurricane produced 307 mm in 24 hours.2204 .3360 50 .f)l.1633 .~49~ . Even the 10-year value was increased substantially by this one event.0998 .2335 .1642 .3315 .1742 .1700 .4060 .\l. lC.2193 .3431 .2229 .1731 .1652 .4210 .2474 . the 100-year value was found to be 155 mm.4234 .3803 .3628 .i60i .1690 .1026 .3150 .EM II1O-2-I4I5 5 Mar 93 11) ’l)nOl. U.3978 .4032 100 .1670 .4773 .i616 .2293 .0980 .2241 .1833 .4411 .2217 . 105:) .0943 .3209 .4384 .4282 95 96 97 98 99 . a 40 per cent increase.S.104s 1(142 .0954 .3271 .3349 .3250 . The maximum event during this period was 171) mm.3456 .0919 100 .io~~ .~409 .2959 .4494 .1764 .4332 .3293 .3780 . The computation of 100-year 24-hour rainfall based on the 51 years of record through 1955 resulted in a new estimate of 218 mm.4307 .4522 .4706 73 76 77 78 79 10[.4005 .3691 .3900 .lalil ~~ . .2379 .2280 .2 .3131 .2307 .3383 .3229 .lW .473!) .4258 .3757 .2254 .3094 .3041 .2441 .3734 .3024 . 10C4 .1599 .4148 .2321 .2349 .1823 .3875 85 86 87 88 89 .4466 .3712 . c-lo .1710 .5 (continued) Return period (years) 11 .io35 .3850 .3007 .4643 .480: .0973 .3588 .1798 .3i13 .2457 .4612 .) 5 .3951 .i624 .1720 .74 .~~:~ .4(.0959 .3925 .4358 .3076 .1010 .1775 .3608 .1848 .2991 .4089 . Based on the 50 years of record through 1954.4581 .0986 . Connecticut.0992 .3826 .4438 .3189 .3i69 .

366 probability A. is equal to ‘rn’l=(n!(~:n)!) pn(i-p)Ln where p = I/T.30 liYDROLOGICAL ANALYSIS It may cipitation of times. The phenomenon happen.185 0.EM 1110-2-1415 5 Mar 93 5. = 100 years.3. that t = T.003 is of P~.061 event 0. prethe magnitude P > PTr does not occur at all. probabilities for various values of n are: n Prufl~ The discussed overall in kc. Assuming. for example. i 0. or greater in t years C-n . or that it occurs several probability that. during a given period of t years.8) the o 0. a respective will occur n times. however. then (5.7.370 2 3 4 5 0.015 occurring 0. years. that during a definite period of T.. 5.

6-3a. weighting with the generalized skew. 2. as illustrated in Figure 6-2. “N” events from the systematic record are given a weight of (H-Z)/(N+L) 01]the assumption that their distribution is representative of the (H-Z) remaining years of the historically longer period. The following example illustrates the steps in application of the It does not include the final step of The historically adjusted skew developed historic peak a~”ustment only. individual flood events should also be plotted for comparison. in which the historically adjusted order number of each event “i” is computed from equations 6-6 and 6-7. by this procedure is appropriate to use in developing a generalized skew. In such a situation. they can be adjusted to give the equivalent historically adjusted values using equations 6-1. The computations can be done directly by applying the weights to each individual year’s data using equations 6-1. The historically adjusted frequency curve is sketched on logarithmicThe probabiiity paper through points established by use of equation 6-5. Historic knowledge is used to define the historically longer period The number “Z” of events that are known to be the largest in The remaining of “H” years. the follow- ing analytical techniques are used to compute a historically adjusted logPearson Type 111 frequency curve. 6-2b. the historically longer period “H” are given a weight of 1. 1. 6-2a. If statistics have been previously computed for the current continuous record.0.EM 1110-2-1415 5 Mar 93 APPENDIX D HISTORIC DATA1 Flood information outside that in the systematic record can often be used to extend the record of the largest events to a historic period much longer than that of the systematic record. 3. The histor- ically adjusted plotting positions for the individual flood events are computed by use of equation 6-8. 4. 1 Reproduction of Appendix 6 of Bulletin D-1 17B. Figure 6-1 is an example of this procedure in which there are 44 years of systematic record and the 1897. and the completed plotting is shown in Figure 6-3. . 6-3b. The com- putations are illustrated in Figures 6-1 and 6-2. and 6-4b. and 6-4a. 1919 and 1927 floods are known to be the three largest floods in the 77year period 1897 to 1973.

for Beard formula. Appendix 5) nunber of historic peaks including high outliers that have historic information number of years in historic period number of low values to be excluded. The event numbers “E” will-range from-l to (Z+N)O logarithmic magnitude of systematic peaks excluding zero flood events. nunber of incomplete record years (below measurable base). a = 0. high or low out’ iers logarithmic magnitude of a historic peak ‘ncluding a high outlier that has historic information number of X’s mean of X’s historically adjusted mean x Xz N M fi d m s 3 G E K historicallyadjusted order number of each event for use in formulas to compute the plotting position on probability paper standard deviation of the X’s historicallyadjusted standard deviation skew coefficientof the X’s historicallyadjusted skew coefficient Pearson Type III coordinate expressed in number of standard deviations from the mean for a specified recurrence interval or percent chance computed flood flow for a selected recurrence interval or percent chance plotting position in percent probability that any peak will exceed the truncation level (used in step 1. a = 0. peaks below base. For Weibull formula. and low outliers which have been identified constant that is characteristic of a given plotting position formula.5 systematic record weight D-2 Q ?P ? z H L a w . a = O.3. such as: number of zeros. and for Hazen formula.EM 1110-2-1415 5 Mar 93 E DEFINITION OF SYMBOLS event number when events are ranked in order fron qreatest maanitude to smallest magnitude.

when:l~E~Z . when: (Z+l):E:(Z+N+L) Iil =WE (6-7) (6-8) D-3 .2)S3G N + UN +3W (N.l)s2+WN(M-~)2+z(Xz (H-WL-1) H-WL (H-ML-1) (H-wL-2)~ 3[ W(N .i)2 (6-3b) s= -2 W(N .i)2+ l(XZ .i)’] (H-wL- WNM +~Xz H-WL 2)1 7 & .1) (M-~)S2 (M - i)3 +X(XZ -i)3 (~-Ab) 1 (6-5) (6-6) Log Q=~+ti i= E.i)’+x(x . 3 1 (6-4a) (6-2b) .i)2 (H-wL-1) (6-3a) H-WL ~H-wL-1) [WI(X .EM 1110-2-1415 5 Mar 93 M= -2 L H-ML (6-2a) s= Wx(x .(W-1) (Z+ 0.5).1) (N .

aoo 10.79 65.82 24.14021 o.10 70.200 3.21 66.0SS40 -0.s2941 3. 1930-1973 (47p) mtipnod: la97-1973 (77p) 14-44 .100 9.94 71.S2 93.23 443s 46.0.M 33.s2366 3.-0 0.IW7 .32 69. ?i-77 ~ Yur Q (c6) -Y =x mpuun k -* w-w Evau N=E W*OAr PlOtt~ mltlm N&r (Welbull) PP x-(x-n) 1897 1919 1927 1935 1937 1946 1972 1956 1942 1950 1930 19s7 1932 1973 13s2 19s5 1936 1948 1939 1945 1934 195.56 3. W. DA 205quareda -Id: 1897.U716 3.31 74.76 51.06 42.004S2 3.67S60 s.s6212 0.00 9.76 37.19467 420796 -0= .45 33.43 16.s9021 3.50766 3.03 U.z-3.320 3.W 66.om as .73 78.36327 0.16726 0.ss182 1 2 3 4 6 6 7 8 9 10 11 12 13 14 15 16 17 1s l.920 1.64 4S.2S717 4.M 14.s40 5.11265 0.661s2 D-4 .62 63.47 61.000 13.330 -m 1.49060 -0.s5612 3.7s0 3.711s1 3.50 20.71 9.4s656 3.17740 0.1399s 4.04003 4.5513s 0.12 Mm 56A9 66.m -0.6s 12.36327 0.U 63.76051 3.76746 3.5s0 “ 1.99 76.23045 4.42251 4.491s6 3.6* 3.000 4.71 46. 1927.17 57.96475 3.640 7.740 4.57749 3.1s0 6.S40 5.32222 4.73670 3.1s435 3.00 1.57 3la 32.00 1.0576s 0.lm 4.6304s 3.oso82 0.230 5.s00 17.01173 0.94 34.35607 0.7!z154 3.51464 0.69 50.07918 4.6666s 3.s6 87.027-o4.2s 2.60 19.95713 3.0.460 6.004W 4.02 7.62662 .21 27.51 89.m 2.07918 -o.660 9.26 72.67 4. 1919.07 1275 14.071s6 0.52 36.01 66.42407 0.96904 3.100 9.16 22.91 40.EM 1110-2-1415 5 Mar 93 ~m F] Station: 3.2A323 o.0.27s95 0.85 5.002S6 JI.14 98.00 1. Bigsu16yhst Bruaton.7066s 3.95 6s.6055.62 73.160 6.220 3.s3s49 0.17 69.26661 0.77379 3.740 2.66867 3.740 6.460 7.65 63.270 442W 3.6S 61.241s2 0.03470 o.19 16.60 76.39 46.43775 3.OSO 2s00 2.ss 41.56146 .S8 39.74s63 3.400 Zoso 1.s0 37.071s6 4.000 21.s7634 3.15s09 0.s9 %.800 12.39794 4.42 74.630 5.00685 4.53 26.34 6.13 29.07 49. 1944 1951 1967 1971 1963 1949 1970 19ss 1952 1947 1943 1961 1966 1964 1933 1964 1666 1%9 1%3 1959 1621 1666 1640 16s0 . 30 31 32 33 24 26 36 27 36 39 40 41 42 43 44 45 46 47 I__ 1.67 22.07918 4.89 29.11663 4.W 5.66 7.21267 33-1 3.23 31.01684 4.620 7.48 21.29 ] m m m m & G v ~ m e & II z G g : 21 2! 23 24 26 M 27 28 29 4.69221 3.62114 3.71s50 3.100 3.000 18.S7578 3.21 64.060 7.o.9s0 3.40194 .56 66s 69.35 43.4s0 1.60 35.941 25.67 91.04 24.72 9.770 3.130 5.s6206 S.97 27.00 4.12 17.35 18.5s0 6.S5 61.1s -0.07s32 1.000 12000 Il.53 63.39 11.

738 2. 6-S~ z GM = 99 95 :.00019 0.(0.02413) (0.0418) (0.37648 UXX3 = -0.08455 3.024. 6-8} a = O.215 2.00409 .391 40.302 q 0.59289 0.38723 4.64599 -1 .47566 -0.33 10[-0”326M [Eq.09755 U>. .07374 Solving (Eu.1) = 0.U1 F=E=l fi=E=2 tii=E=3 (Eg.28110 1 .78808 (Eq.200 9.67313 -0. For E=3.67142 0.32340 3. 6-4d) (77) c= (76) (75) (0.34657 B = 286.190 16.11751 Xxz = 3.30868 4.94tUEL = Solving G= w (Eq.387 (3.71581 :2 ~ = 6.5) For Ueibull: D-5 .11751/77 = 3.13018 ZXZ’ 12.50551 0. Solving (Ea.22132 4.07485) =M — 1X3 = -0.646 20.00067 0.71054 1.08602 4.00409 -0.! .o“929’3[o-8864911 Page 12~ + 0.0418) = -0.@351 33 .z = 5. 0.ANhLjAL PEAKs (Contlnuedj Solvlny (E~.EM 1110-2-1415 5 Mar 93 Figure 6-1.71600 3.?4929 2.26 (0. FP = (100) (~)/(7b) (2+1)=4 (z+ N) =47 For45ES 47: == iii= (1.682) (E) 2.04269 3.08 0. For E=2.01 Solving 2=3 For E=l.07074 0. 6-3d) xx = 162.37046 -C.475 61 .103 1.05159 2.5.70802 0.71581 3.24315 o. 6.95907 4.302 (0.100 12.07227 Solving ii+(g (K)= Log Q M = 3.355 24.47266 3.37021 0.969 5.34535 3.2) .281% -0. Page 13~ 77 + 0. 6-71 (ft u 3/s) -2.32934 -1. HISTORICALLYii[l Gtl TIO LOG PIA~SOh-TYPE 11: .32666 = 0.28898 .34657/(77 = 0.~ Solving (Eq.98733 286. 6-61 -:. 50 20 10 4 2 .13705 6.60719 4.682) (1.24014 3.24326 0.04!8) = -0. 9.3 Sol ving (Eq.84141 0.1503251 = 0.63317 Zx: = 0.89138 1.07074(-0.84180 1.92913 = lo[-l A = -0.07485 Solvinq N= B=0.20952 Zx: = 1.682) (E) .40155 U~X = 273.

76 :3 .69094 s = standard deviation of logarithms = 0.46531 I 0. HISTORICALLY WEIGHTED LOG-pEARSOi4 TYPE 111 .0000154) + (0.ANNUAL PEAKS Results of Standard Computation for the Current Continuous Record Big Sandy River at Bruceton.07140) + (1.36774 I 0.69094) + (12.2 ~= 0.30400 I I (Xz.13705) = o 0835.18746 Adjustment to Historically Weiqhted 77 Years Historic Peaks (Z = 3 Years) Year 1897 1919 1927 1 Yz (ft3/s) 1 I I I 25.18746) + (3)(1.98733 Z=3 1.31740 0. 6-2b): I.07140 = 0.13705 H = 77 0.0000154 52 = (1.02413 )[ (1.07140) + (1.60641 I 0.000 21.02487.28898 = 0.16762 Summation N = 44 Solving (Eq. I 0.01908) (-0.02487)(0.000619) + (1.22300 0.98733) 77 = 3.6(3182)(44)(0.000619.000 18. . 6-1~: 12. TN.68182)(43)(-0. (Xz.50g I I I I I Log Yz = Xz ‘Xz .0418 D-6 .26717 iO.02413 Solving (Eq.J Solvinq (Eq.71581 Solvinq (Eq. (M-fi)3 = -0.X) = -0.7Q802 W = (77-3)/44 = 1.68182) (44) (3.26721 ~2 ~3 = 0.01908 G = coefficient of skewness (logs) =-0.205 square miles (44 years ) = number of observations used = 44 = mean of logarithms = 3.68182) (43) (42) 44 (0.68213 0.68182 1.68182)(44)(-0.1)3 I 1 .X)2 = 0.68182)(43)(0.70802)] 0.R 4. 6-4b): c= 5= 76 (717 (0.E)2 I 0. (M .83990 1.55136 I . 6-3b~: (M . #3-6065 N M DA .EM I11O-2-1415 5 Mar 93 Fiqure 6-2.08351 = 0.39794 4.32222 4.

1919 and1927) plotted as largest in77 year period (1897-1973). . — l+.— -— — . I lo 0.— —— . Recorded peaks (44 years. .— z 4JB ‘- ANNUALPEAKS HIS1ORICA11YWIIGHTIO 10G 4’ ‘ t- —.— . . i: i.000f)4 ——.. O.i ( BIG SANDYRIVIR Al BRUCEION. TN : —-. i . . .. I _____ _ < .71581 ?=0. — —.. . / 1. .__ .— PEAR$~N TYPE III FIGUR[ 6-3 —- — —. . _. / . i..I I 1 I I I I I 1 1 .—.58 years in the longer 77 year period. -. .-.— —... Points plotted by Weibull Plotting Position formula.. “1:”” 7--4——.. . —— — —-. —— . G= Gw=. —..O4 H=77 &3..— —— ——q — 2=3.--1“J.. . .— ...-— l-: r-–7-r. r .-— -. I ._ 0 q EXPLANATION Historical peaks (1897. 19301973) plotted — ———. Weighted Log Pearson Type Ill comp~ed curve._ —. . / . —. N=44.289 / such that each point represents 1.

what is the probability of obtaining an X greater than 3.2% chance that a sample with 18 degrees of freedom will have an X of 3.117 30. From a table (Appendix F-5) for the Chi-square distribution with 19 degrees of freedom: Exceedance Probability 0.128 <0< <u< 19(14. ~nd ard Deviation Find the 90% confidence interval for the population variance from a sample for which S2 = 14.019) From a table for the t-distribution (Appe~dix F-4).197 E-1 .EM 1110-2-1415 5 Mar 93 APPENDIX E EXAMPLES OF RELIABILITY TESTS FOR THE MEAN AND STANDARD DEVIATION Analysis of 19 annual flood peaks indicates that ~ = 3.4918) 30.7782.95 0.4681 )2/19)% Prob (~ > 3.8876? For N = 19 t ~ = 3. are distributed like the t-distribution.8876 S = 0.117 27. Sample values of variance (standard deviation squared) are distributed like the Chi-square (X2) distribution.144 9.7782 K-p = (S2/N)K 3.8876 and S = 0. It is hypothesized on the basis of a regional analysis that the true mean is 3.8876) = Prob (t > 1.14818) 10.4681.4681 ~ = 3.8876 or more.4818 and N = 20. If ~ = 3.7782.144 The 90% confidence interval is computed u (N1)S2 <u< (N-1 ) S2 x. where the population variance is unknown. there is a 16.05 Chi-sauar? 10. Sample means. x& 19(14.7782 = ((0.8876-3.

EM 1110-2-1415 5 Mar 93 APPENDIX F STATISTICAL TABLES F-1 .

16 27.43 94..57 34.49 30.23 22.19 5.13 63.79 13.70 94.03 79. .34 94.17 53.66 32.73 57.66 71.46 18.90 53.40 79.97 77.32 25.10 41.81 58. Number of Values in Array(N) 4 15.69 45.59 43.60 35.68 21.22 19.58 32.68 22.43 30.14 35.00 48.05 68.53 87.47 64.41 78.00 1.92 35.25 86.33 90.13 7.. .55 58.81 3.09 30.49 91.81 93.83 52.65 24.95 54.46 67.40 96.35 76.88 84.56 2Q..73 85.10 93.68 88.68 59.30 73.46 44.86 71.11 93.93 91.91 31.36 96.95 70.41 70. .70 7.49 89.67 90.32 52 2 29.08 88.65 1.59 17.37 98. .87 29.90 1.15 66. .33 36.16 48.66 86.72 40 (m 1 z 3 4 5 6 7 8 9 10 11 12 13 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 I (m) Order 7.06 63.69 65.46 85.70 69.09 1.82 6.92 58.41 18.48 5. 1 50.83 10.06 91.05 5.90 18.18 78.12 33.02 58. .90 39.24 75.37 28.30 98.62 96.51 65.30 20.90 60.44 41.59 26.62 58.57 3.04 89.66 33.52 27.84 53.32 6.68 96.80 45.85 26..19 1.75 12.23 55.43 1.91 38.55 11.86 36.21 32.10 76.03 17.32 62.56 44 6.50 62.91 7.21 81.32 95.66 75.55 55.35 51 3 20.00 98.95 24.77 9.14 3.13 74. .48 24.90 75.43 92.92 46.22 52.40 80.37 30.91 26.23 33. .27 13.41 52.01 72.14 5. .46 68.09 98.43 48 6 10.60 26.65 71.00 93.17 33.95 27.57 84.53 19.61 42.70 23.21 90. .40 49 12. .19 23.41 39..82 37.82 73.53 50.57 20.18 93.47 96.82 1.19 12.79 91.32 73.55 92.27 70.04 62.88 62.95 26.30 1.09 21.62 61.78 67. .39 53.73 1.21 21.79 74.95 36.EM 1110-2-1415 5 Mar 93 Table MEDIAN F-l PLOTTING POSITIONS Plotting Positions in Percent Chance Exceedance order ~ m 1 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 38 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 6 7 6 5 4 3 2 1 (m) .43 74.83 73.17 64.65 92.91 54.94 72.56 39.68 41 13 5.75 27.77 72.14 24.00 79.90 44.66 20.60 69.50 96.24 14.76 11.41 60.77 62.41 69.59 56.00 21.76 23.04 89. ..71 39.21 71.23 51.17 42.56 21.08 12.61 13.27 94.68 54.43 32.77 52.85 66.38 51.63 67.31 29.61 54..20 50.36 33.16 71.96 50.47 76.99 33. .74 39.39 43.14 45.87 46.34 37.97 85.50 1.51 6.48 36.91 64. .29 88.25 25.85 88.50 7.59 98.37 59.33 15.61 35.75 29.15 15.34 24.32 20.6.78 10.57 98.08 48.38 50 .10 20.39 98.17 62.25 54.23 24.41 78.06 2t3.77 71.43 75.20 64.23 22.45 40.09 98.53 9.11 39.96 5.68 75.57 63.37 90.76 78. .77 73.00 47.07 31.71 6.43 87.35 94.18 86.92 6.04 43.14 79.00 58.81 41.77 40.70 22.64 76.92 66. .01 28.95 14.68 29.45 89.24 72.74 68.41 81.18 57.11 50.1o 68.62 9.00 47.30 50.23 64.03 11.66 60.60 43 Order.75 40.35 50.96 52.64 94.85 35.62 61.32 74.88 37.42 50.89 45.78 29.64 70.87 8.96 7.47 86.54 96.63 5.32 66.36 4.70 86.91 93.76 46.78 34.59 77.43 28.59 35.81 98.36 84.35 16.65 19.07 3.12 10.22 66. .01 47. .90 83.65 79.36 25.73 25.87 87.19 92.92 82.63 53. .17 22.77 94. F-2 .79 5.22 85.15 60.77 31.75 69.94 31.63 81.29 81.54 51.37 42.46 47 7 9.78 38.97 15.60 15.11 44.94 92.54 28.23 1.33 66.24 83.72 84.02 50.43 11.59 16.38 34.32 41.66 87.64 42 5.Order 10 1 6.07 39.92 14.29 70.57 94.14 58. .50 32.91 73.22 40.13 31.45 84.06 98.02 34.32 94.15 29.08 81.36 1.78 55.98 85.11 70.50 58.06 42.60 96.26 55.38 33.41 16.81 90.79 91.08 59. .77 44.13 44.30 72.83 61.73 15.40 81.64 62.23 86.90 13.05 16.43 13.30 38.82 70.24 17.00 57.95 78.10 51.90 26.76 81.58 74.57 65.89 98.74 16.13 66.38 37.98 36.67 80.b3 22.83 42.16 75.65 96.12 58.13 48.54 77.31 83.75 60. .88 87.67 32.84 68.76 92.67 26.87 79.35 73.10 35.01 18.42 63.58 13.69 12.56 45.23 43.17 84.66 50.19 36.86 37.64 61.68 21.12 31. .88 84. Number of Value sin (N) .27 36.64 19.83 33..73 82.07 68.17 84.89 51.34 85.52 77.19 77.67 41.43 61.76 20.28 79.64 23.33 62.11 14.66 57.57 83.13 12.59 25.51 4.30 28.99 49. .97 13. .03 48.70 16.50 94.33 54.09 63.87 83.29 43.53 45 Arrsv .32 36.05 47.67 56.35 3.04 17.93 41. .13 27.70 57.10 87.18 25.21 23.16 42.90 64.57 66.34 8. .4b 96. .00 63.85 78. .30 88.30 57.99 66.51 18.17 13.52 71.28 95.83 11.04 68.81 64.63 67.78 43.02 45.63 36.34 6.87 64.52 59.63 50.45 37.93 06.86 92.70 36.78 76.05 46.67 4.44 56.03 69.09 40.57 96.95 91.59 14.14 57.86 59.05 70.96 54.95 50.21 3.05 90.00 47.13 87.63 65.96 77. .22 50.05 26. .45 3.00 60.00 68.65 22.47 50.70 98.22 56.94 3.38 85. .05 67. .56 76.36 19.04 55.57 45.89 23.30 9.50 37.22 61.10 30.39 41.33 5.54 10.05 20.71 9.07 8.57 14.15 16. .00 47.63 17.97 25.80 31.68 3.50 46 e 8.24 3.82 46.26 11.68 93.01 61.87 20.50 16.86 89.39 60.25 68.55 42.69 55.43 12.15 6.26 26.27 79.08 9.59 14.95 43..50 63.50 79.02 9.03 83.70 45.05 52.85 79.25 83.68 89.08 37.43 80.19 54.24 46.12 7.19 53.37 78.97 46.98 45.07 54. .42 21.64 91.86 16.93 56.34 35.57 6.10 56.96 28.96 67.32 7.94 35.44 66.10 82..99 76.49 77.50 33.79 72. . .52 90. .00 47.57 82.98 60.05 41.95 74.05 80.95 49. .27 10.32 92.71 51.79 24.35 71.10 11.42 82.25 15.62 44.61 52.57 .00 73.44 74.17 27.81 18.83 31..48 50.50 88.34 20.

95 76.11 62.25 50.07 30.55 62.88 14.66 44.05 70.16 32.16 90.04 95..55 9.07 5.32 5.25 7.99 29.19 2.39 26.33 44.45 28 26 2.76 20.42 76.35 45.63 43.67 10.73 38.62 88.50 67.31 34.02 27.oo 43.81 36 .02 55.98 63.51 53.22 96.86 37 19 3. .37 53.72 41.97 74.07 82.60 17.72 66.62 55.00 54.58 17.00 69.05 82. .95 11.85 47.95 53.43 81.92 12.72 54.96 95.35 83.37 34.88 7.74 42. 17 4.55 45. .08 52.25 20.95 57.65 45.79 90.47 92.55 52.17 65.53 16.54 46.23 28.30 35.90 15.86 79.59 85.53 39.51 37.56 79.49 40.28 51.72 50.42 10. .58 46.84 85.23 54.14 18.85 70.85 79.49 5.77 32.28 22.26 96.95 11.39 68.63 45.83 4.83 86.07 2.16 66. .99 83.61 60.46 64.61 57. .86 38.17 96.35 77. . . .97 7.39 88.33 61.84 74.55 94.31 67.49 72.83 11.83 87.96 70.01 63.17 89.69 73.22 14.25 27.54 13. .61 4.82 16.65 34.22 73.12 65.74 10.67 34. .46 45.83 67.56 13.30 69.00 55.15 24.72 59.81 53.30 46.90 65.98 90.38 39.85 87.70 58.88 56.52 58.05 24.40 31.20 66. .32 22.04 20.05 59.72 94. .81 38 .00 17.21 92. .98 63.78 18.00 76.06 8.05 17.54 83.26 34.51 85.81 61.65 52.23 62.57 9.44 22.00 53.37 38.51 21. .00 56.07 71.57 12.00 47.62 31.51 2&.33 76.42 79.07 30.35 64.66 1.47 79.55 35.35 74.24 90.31 34. .38 15.93 75.14 13.37 46.16 70.61 69.22 18.92 94..40 4.62 56.08 33 21 3.53 77.26 10.61 14.34 16.16 60.69 60.77 85.91 68.09 95.85 8S.92 9.66 78.49 22.39 93. .42 67.76 97.97 2.46 79.59 69.32 81.54 26.08 76.40 2.48 92.79 21.84 4.79 94.23 12.47 77.24 95.22 27.37 65.15 7.80 20.16 57.56 86. .06 11.75 84.40 15.43 72.76 98. .64 50.60 89.03 90.35 27.28 28.74 13.46 11.15 97.21 56.06 50.70 46.43 96.69 96.35 90.65 39.86 65.08 43.50 87.24 10.10 91. .75 61.42 82.48 35.12 46.36 70.45 31.96 61.50 63.24 73.19 37.58 5.17 44.72 58.37 29.25 73.76 39 15 4.44 80.90 37.80 10.03 63.07 19.93 64.15 65.23 8.78 1.86 25. .48 55.79 93.47 93.67 75.30 56. S8 61.43 84.35 41. 14 4.68 59.40 50.53 53.10 26.93 32.76 2.88 61.68 88.12 56.22 95.30 53.55 81.00 96.63 39.&l 8.25 23.19 50.93 26.39 72.21 4.27 97.86 94.02 4. .81 90.85 46.78 9.78 19.57 2.75 43.68 59.84 82.43 26.10 64.78 91.70 61.60 25. Numbaraf 18 3.45 57.63 34.97 25.19 39.82 65.91 85.51 82.18 42.37 97.23 85. .35 42.70 24.75 97.63 77.47 70.08 13.25 71.85 6. Order 23 2. .25 64.03 74.05 63.14 69.43 29.63 75.62 7.28 52.82 70.67 50.48 98.31 13.73 25.84 23. .74 28.26 77.14 32 .90 19.89 32.sluesin Arrsy (N) .17 41.05 30. .81 68.15 19.89 55.60 76.74 58.13 61.46 83.47 63.61 67.26 67.33 18.47 60.40 92.49 88.23 2.05 97.97 12.28 30 24 2.63 78.64 94.04 17.88 43.17 72.11 73.60 60.00 50.94 57.75 40. .96 35 Valuss in Array(N) 20 3.03 16.77 38.05 35.38 87.24 38.EM 1110-2-1415 5 Mar 93 Table F-l (Cent) MEDIAN PLOTTING POSJTIONS Plotting Positions in Percent Chance Exceedance Order (m) .5B 25.13 40.60 91.63 51.58 26.33 97.61 22.06 41.32 50. .55 76.50 90.97 92.83 37.21 86. .61 25.15 60.05 15.41 8.81 55. .83 2.74 40.42 67.44 37.38 75.53 27 (m) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 (m) Order 1 2 3 4 5 6 7 8 9 10 11 12 13 lb 39 36 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 18 18 17 16 15 14 13 12 11 10 9 6 7 6 5 4 3 2 1 (m) Order 16 4.52 11.91 81.37 74.07 24.37 60.03 29.84 47.82 42.00 47.85 62. . . .87 54.05 14.21 31 . .95 79.09 25.08 96.26 75.09 24.45 45.67 33.55 1S.68 71.52 38. .56 69.45 12.96 38.27 54.79 17.58 8. .99 53.90 97.77 8.83 50.57 81.76 48.06 16.46 27.68 32.39 24.90 34. .59 92.36 76.50 20.76 69.58 97.48 9.37 30.97 36.00 44.00 47. .16 95.23 56.66 86.73 6.19 36.65 48..05 56. .56 71. F-3 .01 80.02 32.00 55.83 62.42 72.91 34.87 5.93 64.10 74. .84 6.67 17.07 76.53 30. .77 52.75 38.26 86.51 82.18 22.44 20.44 14.54 50.09 60.49 75..88 60.50 50.50 44.10 79.10 71.21 14.01 31.03 26.94 76.07 66.77 81.11 33.01 30.00 46.65 20. .26 67.48 11.58 20.84 42.78 06.26 64.10 7.12 51.72 71.16 8.16 63.95 52.02 34 22 3.17 5.32 48. .77 82.02 29.95 31.03 80.90 26.53 44.03 66.18 57.11 51. .84 84.90 69.24 44.02 35.35 28.97 64.62 32.25 22.36 29 25 2.14 29.04 23. .35 61.38 6. .74 7.78 66.66 95.28 35.95 36.36 51.01 36.75 74.37 5S.96 73.41 41.52 15.74 83.01 41.81 51.00 46.81 43.10 84.59 S7.70 39.42 88.00 47.88 27.45 89.69 27.45 84.10 23.37 18.65 12.54 66. .95 86.35 52.75 72.47 16.64 37.00 89.89 21.95 93.14 14.45 46.63 33.75 84.57 79.62 77.24 11.50 21.00 54.39 32.08 85.94 51.63 70.62 36.70 35.34 92.00 9.25 96.83 12.95 26.01 9.99 82.00 32.77 65.15 30.73 87.87 67.35 68.36 37.13 87.56 14.99 95.86 77.65 80.93 81.63 27.37 42.32 65.93 23.86 76.95 64.67 33.63 40. Number of V.05 81.75 15.60 43.63 6.65 50.12 87.21 29.63 74.70 73.00 46.21 75.53 1.44 7.79 36.74 15.97 40.27 21.27 88.64 47.19 95.30 1.30 36.42 97.28 38.18 31.30 47.51 86.70 93.17 82.45 72.41 1.57 21.56 56.18 25.28 58.65 10.33 56.01 17. . .64 23.65 41.65 20.79 80. .03 16.12 23.40 68.36 62.03 97.99 55.

22336 5.38961 2. 0.- 3.~5 ‘1.84981 .09~5 0. .03443 3.84389-0.8L864 1.1 2.93183 4. F-4 .07126 3.14712-0.61539 2.39554 0.33869 2.7 0.05138 4.82798 4.10697 2.38652 -0.79421 1.3 0. .33204 1.82359 2.23353 -0.51768 1. mla -1.79025 4.91492 -1.70543 4. .*3 o.53663 1.- 2.M5 2.31760 1. .33 2.810~ 3. .32263 4.3 2.31084 2.95472 -1.75514 2.55304 5.9 -1.82629 6.06651 -1.264U 2. O361O -0.90973 4.96959 6.64162 0.02933 -1.33~ -0.2542? 0.43839 4. .8$573 2.02328 3.09320 3.66023-0.33684-1.531U -0.84422 -0.67279 1.75221-0.4~ -0.31671 1.0 80.32014 6.=18 6.94871 4. .78323 5.o 20.73257 1.29517 0. .32835 -2.22422 2.0 1.08711-1. s7 4.50701 4.M -0. Bulletin 17B.=139 3.95%7 4.=1 3.13232 3.79701 1.1~7 -1.0 90.99016 4.6 1.-2 1.084234.55295 4.66651 5.61594 -1.89101 .--0.06302 -O.0.15935 2.87227 2.49101 -1.64191 ‘1. . .65176 5.10394 ‘2.35062 0.--0.~72 2.28387 -1.21618 -0.4 0.8W32 43.60872 3.24313-1.84848 3.64262 0.92868-0.71067 4.09323 3.76482-0.37087 -0.65285 -1.94181 3.18627-1.21461 5.65508-0.1.64335 1.99744 3.95735 3.4 1.~76 1.0~1 -1.48737 3.80W 6.3= 0. 2.82516-0.= 1.7 2.27365 2.87816 2.74.0 m. .74708 -0.EM 1110-2-1415 5 Mar 93 Tabie F-2 DEVIATE SFOR PEARSON TYPE-111 DISTRIBUTION Positive Skew *W tifzicimt (G) 0.47S34 1.W28 -o.667= -0.lZ169 3.55527 -1.4 2.77~ 0.10768 5.30936 1.97026 3.10022 4.35289-1.77X9 2. .31243 -1.51467 5. .s1808 -1.443W -0.92216 6.68971 1.72935 l.~522 -0.27037-0.37703 3.79472-0.30685 0.73860-0.77686-0.340s2 1.77428 1.90776 4.72226-0.37186 5.O3.76676-0.-O 1.0.52257 -1.87301 4.3=1 2.44438 4.018S 4.38807 -2.84540 4.12398 3.96233 2.29832 5.63223-0.- -0.23768 2.-1.48672 1.= -1.20562 4.88381-0.42040 1.83E?7-0.67532 1.7Z7E 0.42426 6.49811 2.W570 +.78838-O.7?082 ~. .7 1.72957 3.-1.29377 2.0860E-0.75048 1. .71635 5.0 0.2 0.52694 6.3.42292 6.8K175-0.8 0.21.=0 -0.0~ -0.68971 5.74497 4.~178 1.67191 0..1 3. -7 -0.23322 3.m95 2.70208-0.00335 3.72422-0.6803 1.18574 -0.02973-0.80837-0.61720-0.85161-0.M1l -l.34039 1.77964 3.24371 3. .20020-0.0 2. XI.5 0.14700 4. *at-o 99.61472 3.0 95.74049-0.01S62 -0.00000 -0.82916 l.0.86a08 2.36993 0.28350 0.85636 2.4W5 -1.844S -0.3MM ‘1.25039 2.2 2.5 2.95392.40670 2.0 10.=73 4.12356 4.94900 3.02330 3.84669 6. .79950 0. .9 4.3=5 2.17968 0.58393 5.6 1.64485 1.‘0.68572 2.62377 0.54814 -0.1 0.0 -O.75347 4. .84611 -1.38804 3.75752 1.34047 l.85426 2.44942-1.736104.5 0.29626 4.33035 3.57883 1.0 2.=7 5.08E60 3.72562 1.63841-0.40109 3.68532-0.4.88029 -1. . .32665 l.~506 7.--0.2 2. .06926 4.71435-0.60944 l.=5 Source: Appendix 3..67344 2.23075‘1.14944 3.63044 0. .54421 2. .- .0071.36267-1.54206 3. reference (46).86959-0.66603-0.3010S 1.8~ -0.31872 0.20578-1.lon 0.73271 -le66M1 -1.67683 2.81638 O.79302 0.36566 3.25256 -2.83639 0.58480 3.4= 1.01177 3.76321 2.99978 3.64M -1.1.70~ 3.17640 -2. .42345-1.95063 2.6 2.0~ 6.0~ 3.1 1.327624. .92202 3.47Z 2.64609-0.88995-0.18347 -1.3.95549 2.62a82 1.85703-0.98124 2.38353 0.5 1.57583 2.00803 2.04787 3.9 .6 2.27324 3. 0.64485 -1.22040 ‘1.74325 3.03ffi5 5.04~ -0.ml 3.13578 -O.46674 4. .14042-1.23033 2.0 2. .25824 -0.72495-0.70522 1.0 b~m 5. .00326 2.-1 -1.23114 -0.36655-1. .65718 -1.3 1.-0.63386-0.32S00 1.74537 1.77875 5.86231-0.-8 3.2 1.-1.0~ 4.79668-0.55549 1.58838-1.76242+.04102 3.28M5 -1.91505 -0.31054 1.68256 *.60517 4.16616 -0. .3033.21081 2.32308 1.03247 3. .62631 3.15193 4.26240 2.24439 4.66657 3. . -5 0.64876 0.0. .91456-0.57530 4.13254 7.1- 2.28~ 1.59383 1.78802 1.6 0.44035 1.45762 -1.0 1.- 1.05375 2.a3895 -0.04144-0.33330 1.24516-0.-4.M~ 0.94607-0.37840 4.

02211 1.301O5-0.3 1.95131 0.32800-0.67532 0.30W -2.88075 0.77868 0.W570 -3.-7 0.66532 0.08302 0.77428-1.4 -0.13S78 0.0 .6 -0.7S95410.94945 0.14732 1.94W7 0.68572-1.90.1 -0.23998 o.70543 -2.31054 -2.4- 0.34047-0.0.85161 -1.79002 0.23766-0.43529 1.86e63 0.04144 1. .45762 1.um 0.83283 0.22040 0.90W9 0.EM 1110-2-1415 5 Mar 93 Table F-2 (Cent) DEVIATES FOR PEARSON TYPE-111 DISTRIB UTION Negative Skew 9tal Cimmciti (G) 0. reference (46).75221 0. .14%4 -1.= 1. .=90 1.98213-1.55527 1.-1 2. oom-l.42040 0.=23 0.64485 2.00335-1.30685 0.82472-1.=18 1.00128 -3.8 -1.53683 0. .00000 0.87816 3.51768 -1.6332E 0. m3 0.33904-o. .76878 o. .04993 0.83320 0.20092 2.66964 0.04427 1.48216 1. =878 “0. 76W9 0.79015 0.0 M.32850-0.08711 1.22114 1.17840 2.90992-1.60829 0.- ‘2. .63183 -3.00903 -3.0 ~e 5.01844 2.~028 ‘2.80013 -3.27037 1.70. 7=1 0.28167 1.64305 1.8 -2.58838 1.39S61-1.94498 1.5 -0.5 0.34m -0.29422 2.70209 0.o Fucmt-o 20.131S9 O.3 -2.87683-1. .64335 O.316a4 1.73987 0. 98.39942 2.51808 1. .0.94671 0.63594 1. Bulletin 17B.2431.20576 1.~01 1.68%5 0. E6667 0.85506 1.35062 0.79765 o.61539-1.9 -3.m3.4M39 1.71835 0.66663 -3.zam -o.97301 -4.83222 1.5316 1.99973 2.-3 0.371.6W71 -1.23716 1.67~ 1.84976 0.73610 0.823* ‘1.84611 1. .95083-1.48187 2.53U1 1.e4m 1.03695 1.2 -2.74074 0.17868 o.16974 1.60517-1.81720 -3.49183 1.-5 0.=935 -1.75752 0.74062 1.01810 1.03325 0.32376-0.=95 1.39408 1.88959 2.31671-0. . .90521 0.8S101-1.31760-0. .1s680 1.-7 0. .74709 0.03228-1.- 0.W29 1.80786 -2.93636-1.~22 -0.32412 1..44015 0.68965 -2.22013 -2.0 .68935 0.4 -1.84162 1.01177 -1.99573 -3.0 0.72495 0.79990 0.0~ -1.3.7 -2. Em7 0. -3.78572 -3.83660 1.09323 -0. . .36852 0.79701‘1.2 -1. .74537 0.0 .39554 0.65741.44942 1. 2.3 -0.16397 0.37981 1.5 -1. .21618 1.30279 1.34039-0.63959 0.92580 2.79994 0.85856‘1.94306 2.051.7 -0.68111 0.38795 2.3 -1.ffi945 0.08808 1.44438-1.72422 0.78482 0.42345 1.32309-0.0 -2. .6 -1.28155-0.oEa360 -1.26780 1.14807 0.1 1.25039-0.81864‘l.74073 0.- 1.54886 l.7142E 0.218 0. .1 -2.-9 0.01739 1.58110 1.77716 1.5 -2.86371 0.32635 2.1.73271 1.-0 0.73257 0.28412 -O. .0 -2.W26 -2.31815 1. =742 0.E63Z 0.95186 0.0 93.10743 1.71.15477 1.0= 0.09749 1.18827 1. .-8 1.77~ -2. .631% 0.o M.EJ3604 1.98977 1.-0.--1 .78902 1.12828 1.65718 1.57383 -1.05631 0.12762 1.= -1.84368 -3.27134-1.w 0.70512 0.99499 0.49101 1.28019 1.35114 1.75706 2.77082 0.0 -1.84244 1.83916-1.63672 2.76242 0. .79973 0.=77 -0.75514-1.84485-1.16534 1.2 -0.2 -1.38991 0. .33330-o.78828 -1.84162 O.=9 -0.63841 -3.25253 2.3= 1. .00710 -1.23075 0.16584 1.81638 0. .33035-1.16347 1.38855 1.22203-1.70.54421-1. = ‘0.14053 -2.68959 0.=178 -0.0.57W 2.68023 0.0.14042 1.7 -1.16884 2.21618 1.04898 1.6 -2.83361 2.38267 1.67279-1.97227-1.38333 0. .7n77 0.0 10.28155 1.o- 0.714= 0.95735-1.3= 0.33865-0.36804-1.-O 0.79472 0.75347-2.29539-0.33640-0.32872 O.63~ 0.10394 2.0 ‘4.08413 1.51741 2.91458 0.26808 0.49872 -1.=5 -3.84066 1.94334 -2.74049 0.76n9 0.27365-0.18006-0. 4.09945 0.47~ -0.74087 0.8 -0.W885 o.24526 1.8= 0.25824 1.05375 2.10825 2.6 =11 1. -3.39867 1.94499 2.58807 1.84540 -3.=7 Source: Appendix 3.71435 -2. . .4m5-l.19317 0. F-5 . .06651 0.0M7 -2.3114 1.E73 0.4 -2.0.83044 0. .9 -3. 0.23805 1.85703 1.97980 0.9 -2.95472 2.1 -1.82277 0.-2.89464 0.7W8 -0.71428 0.10465 1. .02933 2. .29443 0.98324 0.=5 95.89894 -1.7~-l.59263 -1.05701 2.47226-1.83639 0.57695 -3.55295-1. 71W o.M728 1.98381 -3.90854 0.06864 1.0= -2.31275 0.28311 2.84422 O.--1.76922 0.25422 0.37640 0.63668 0.65800-2.46232 1. . 3a759 0.38 -2.-3.23132 1.7~33 1.52357 1.74913 1.01.

0193 .9 .3 :: .29 .59 34.0904 .0291 .01 49.317 . .139 .29 2.81 26.639 .4 2.00279 47.2 .97 35.22 16.06 4.587 .00245 Note - Values have been generated by use of computer routines for the normal distribution.14 15.s9 2.61 17.57 11.77 25.75 3. .00641 . .74 35.50 1.714 .58 20.889 .79 11.07 1.00615 .86 38.09 23.0136 .18 5.00291 Incrmsnts to Kin Colunml .0179 .0126 .6 1.43 25.36 19.94 16.480 .81 5.32 1.20 6.53 7.2 2.2 1.26 4.88 15.298 .12 1.256 .69 20.0390 . .307 .79 1. .01 44.402 .00304 .9 3.68 0.0251 .25 39.280 .118 . . . .09 3.0270 .74 2.20 45.82 4.11 22.042867 .00850 .43 24.00481 .144 .453 .63 40.01 3.494 .00667 .9s0 .85 7.94 32.6 .233 .45 33.33 17.16 4.50 27.5 3.05 17.755 .0260 .820 .00331 .0519 . .264 .10 10.0362 .00425 .0153 .38 6.36 35.00 46.0577 .205 .111 .43 23.0172 .76 23.85 27.657 .94 2.75 3.75 9.336 .85 14.65 8.51 9.0165 .3 1.00544 .05 4.06 6.48 18.226 . .00225 46.56 28.71 4.289 .67 16.107 .2 3.84 3.00884 .11 14.00256 47. .83 31.62 1.09 30.10 24.523 .0736 .0762 .1 2.01078 .554 .00443 .8 2.64 29.93 9.00216 1.83 40.67 2.0 . .08 5.604 .12 25. .0557 .4 3.9 2.44 2. K .01 2.0664 .19 18.46 4.46 22.0142 .90 10.07 2.00784 .60 14.40 44.199 . .93 6.0874 .66 19.27 3.35 11.0404 .114 .41 15.51 21.00501 .0968 .0159 .36 1.28 28.83 1.00462 .0224 .00591 .64 39.240 . .00 29.61 43.01 11.181 .36 3.0419 . .14 11.46 12.70 9.22 1.62 13.169 .47 38.0845 .939 .65 3.01037 .31 9.00360 .85 3.67 2.914 .70 1.95 4.0313 .0936 .17 1.539 .0 1.76 24.104 .427 .7 2.08 .43 5.1 1.0121 .676 .00345 .0598 .776 .30 10.03 8.10 . .44 1.149 .22 2. .6 1.16 3. .0280 .21 43.50 1.37 4.92 10.357 .0501 .09 46.EM 1110-2-1415 5 Mar 93 Table F-3 NORMAL DISTRIBUTION Percent Chance Exceedance for Given Normal Standard Deviate (K) .00375 48.07 38.100 .0 .23 6.80 44.00391 .15 12.46 3.159 .33 1.32 32.0641 .57 31.0302 .466 .842 .0434 .212 .46 30.175 .25 .0711 .51 7.30 5.6 ::: 3.0208 .347 .00408 .1 .798 .0233 .570 . .68 37.379 .62 15.90 37.0325 .66 1. .126 .0619 .01 7.21 5.0 3.35 14.62 41.49 16.06 .164 .0185 .3 3.695 .77 18.0147 .92 3.0131 .07 .59 4.00695 .66 5.0467 .0 2.734 .36 2.00235 46.36 29.04 .46 1.193 .92 1.36 14.39 1.62 2.74 1.0099 .02 42.00522 . .83 34.16 .122 .94 4.29 37.508 . F-6 .71 10.0200 .6 2.248 .0117 .440 .34 7.81 2.87 13.1 3.326 .51 10.22 .0376 .00958 .7 1.00267 48.78 22.23 12.18 7.60 45.0350 .78 6.13 .0216 .38 8.55 3.621 .00567 .00 50.7 .0337 . .154 .92 28.39 1.5 1.46 26.12 9.83 21.86 1.0789 .69 12.68 2.0242 .02 1.13 36.02 49.07 33.9 4. .27 20.0538 .51 2.97 1.49 6.00724 .69 33.05 .14 22.167 .0 5.0483 .00817 .41 42.15 26.00753 .56 6.43 1. .368 .35 6.0687 .64 6.11 13.22 41.0816 .58 1.39 13.19 .04 40.017 .56 2.81 42. .00317 . .415 .131 .21 34.94 5.0112 .866 .90 18.55 5.72 30.272 .964 .0 6. .69 7.52 36.391 . .3 2.03 .92 12.14 2.135 .8 .00920 .54 1.06 19.219 .4 1.96 20.77 17.07967 .20 21.0450 .20 31.072 .28 1.044 .21 27.5 2.

873 .1.255 .= 2.350 2.678 8.539 4.706 3.979 .203 4.372 1.= 1.Qa 1.m . . .264 2.EM 1110-2-1415 5 Mar 93 Table F-4 P ~ R NTA NE-TAILED t-DISTRIBUTION .622 1.326 .319 1.a5 3.lW 3. .423 3.m 2.254 .549 3183.n6 1.354 2.012 2.654 .s?l 2.650 2.%7 2.8% .719 3.767 1.361 2.018 2.623 2.906 .373 3.638 1.762 .104 2.714 4.333 1.505 3.771 1.697 1.4U 1.648 2.X 3.257 .3% 1.=1 . .781 1.352 4.262 2.619M91.616 4.094 4.E64 3<652 3.797 2.76a 1.915 13.598 22.zm 1.364 2.740 1.676 2.163 2.727 .s .649 3.Oofl 4. .303 1.257 .174 ::E 3.452 4.779 2.030 5.84a .073 4.761 4.251 3.467 2.397 1.=7 .8U 5.454 4.657 L59.335 1.665 1.254 .746 1.109 2.435 3.465 2.699 3.09 9.934 2.W6 2.043 1. .622 2.7= 3.= 15. .090 5.964 2.358 2.500 2.299 4.= 1.291 7.055 3.530 .060 4.716 3.= 2.324 4.694 5.046 2.001 from published tables.395 3.52a 2. .848 3. .617 l.576 4.852 3.745 3.261 .533 1.674 3.255 .172 2.018 4.225 2.632 2.761 1.666 .277 .0001 63.241 4.655 .307 3.624 3.24Q 2.173 8.559 .666 3.850 3.m 1.30 .275 4.2!?6 .= 1.224 2.m 2.81S 2.= 1.671 1. .327 3.262 2.381 3.267 0.633 l.333 1. Rcadmce-iu@ ~E 1 2 3 4 5 6 7 8 9 10 11 32 lz 14 15 16 17 18 19 20 22 z 23 24 25 26 z B 30 40 50 60 70 60 so lW 110 z .624 2.959 5.676 1.656 1.20 1.OH 004 7.042 4.549 .m .m 4.= 0 0 3.361 2.3s2 4. .204 4.814 1.254 .318 1.232 3.646 3.303 2.528 4. .717 1.345 1.169 1.257 . .437 4.517 2.- 00 IS.635 3.698 2.038 4.396 3.314 1.489 3.534 .539 2.660 1.326 1. .440 1.870 .063 4.599 *3.886 1.355 3.011 5.309 636. .320 2.231 3. A few values for exceedance probabilities of 0.311 5.W 2.6% 1. .054 .076 1.530 1.552 2.721 4.415 4.642 3.485 3.645 1.310 1.M9 4.610 3.667 1.605 2.m 1.646 3.665 1.205 2.257 . 5.973 1.795 3.461 5.932 1.143 2.648 4.869 9.878 3.442 6.6E 2.076 2.645 . .2% .~2 5.660 1.768 3.678 3. ruu 22.416 3.1.s 4.941 .= 1.078 2.254 4.841 21S u.796 1.4s? 2. .681 2.025 7.563 2.353 Z.479 2. .508 2.457 2.845 2.543 .610 10.W 4.840 2.791 4. .704 2.263 .147 3.539 .m 1.224 3. .393 3.=1 1.s 2.Q7 .258 .235 4.061 2.128 3.764 2.31 . .819 3.m 4. .374 2. .662 1.858 .263 5.147 2. .258 .552 3.001 0.669 1.837 3.m 2.= .962 3.183 3.326 1.314 2.074 6.845 . .328 2.347 4.m 1.005 and less may differ plus or minus 0.m .642 1.951 7.540 8.191 2.7s6 1.7% 1.462 2.453 5.985 4.321 1.703 1.879 .533 .3 4.60 .= 3.= .254 .26s l.337 1.= 3.063 6.256 .719 0.408 3.408 5.965 3.821 2.066 2. .750 3.16.170 3. .845 .000 3.651 . .% .X2 8.273 3.7C6 1.706 1.221 3.849 .965 4.271 .297 4.250 3.mo .396 2.363 1.31.999 2.625 2.% 4.032 5.460 3.014 3.626 2.463 2.221 4. .Za2 .553 . .403 2.025 3.m .m 5.365 3. .4s6 3.518 2.883 3.786 4.S48 l.757 2.530 .144 4.m 6.684 1.s83 3.621 6. .s64 3.323 1.402 3.257 .% .657 .004 1.690 3.7& /u.259 .531 .676 .750 2.701 1.053 10.6U6 2. .136 3.333 2. .787 2.= .% .541 3.659 .326 3.zaz 1.450 3. .* 3. .035 3.0005 0.540 0.878 1. 00 0 005 0.449 2.254 .261 3.311 1.106 3.493 4.m 2.536 .473 2.366 2.323 2.617 ..771 1.034 4.63 3.111 4.635 1.977 2.140 4.82S 1.002 0.602 2.269 l.846 .224 2.137 4.325 1.655 I.756 2.707 3.734 1. ..ls8 3.718 2.267 2.753 1.W1 3.787 3.546 .365 3.327 31.747 3. .826 1.%1 4.733 3.661 2.376 1.294 1.m 31.177 2.564 .639 2.667 1.930 3.644 I.725 3.055 3.664 1.390 4.550 3. m 318.254 .666 .35 .169 3. .531 .587 4.940 2.926 3.2a9 .2n7 3.954 2.341 1. F-7 .869 .676 005 6.234 4.670 2.359 2. .3XI 1.064 2.771 2.325 .= 17.714 1.356 3.435 3.579 3.846 .935 2.358 2.363 3.977 2.= .334 4.916 3.m .527 3.e62 3.832 1.W7 2.363 1.655 .787 1.438 3.I.330 1.slz 1.663 .617 2.721 1.763 2. . .876 S.265 .423 2.533 .542 . .467 3.751 Note - Values have been generated by use of computer routines for the inverse t-distribution.695 4.847 .659 3.773 .061 .729 .551 3.711 1.

575 5.442 53.75 .955 23.825 m.476 33.215 126.342 7. . 379 16.w 29.638 42.571 4.334 From Table 26.6al 76.851 u. Government Printing Office. 443 16.117 10. 3= 26.737 0.3 41.188 26. Washington. .6E6 23.548 20.145 1.102 .91 37.754 70.635 9.749 22.558 3.334 79.229 S.401 42.?= 44.s9 1. 302 59.531 101. W 44.697 39.245 3.675 14.141 28.239 11.336 39.M.565 Ua.485 45.848 14.939 19.860 42.337 42.251 7.= 35.554 .211 .321 128.675 21.2S9 140.605 Z.352 .= 24.146 31.337 21.w7 17.13.037 11. 0.587 28.241 =. . S1 U.455 1.340 12.402 e6. .912 u.645 L2.458 24.165 12.334 69.557 43.070 E.005 .1. National Bureau of Standards Applied Mathemat~al Series 55.8XI loa.718 23. 4= 26. .362 14.289 41.Z7 .952 104. .007 33.588 50.919 16.EM 1110-2-1415 5 Mar 93 Table F-5 VAL U E SO F CH 1-SOUARE DISTRIBUTION .435 37.= 5. 208 149.99 0JIS7 .645 m. Rualmco~ility 0. .145 80.315 b6.865 11.597 12.476 %.379 100.661 W. 037 19.260 8.346 7.064 22.801 34.877 29.348 6.95 .X 2.610 2.314 45.636 56.2s 0. .Q31 7.337 M.790 42. .W5 2e.304 7.898 6.892 %.841 9.181 45.338 16.= 6.lzs 77.320 88. .897 9.841 5. m7 16.679 m.m 40. .317 n4 .459 55.566 38.88 51.860 16.455 4.209 24.240 14.299 29.7W 8.115 . .294 61.675 3.344 17.001 .= 18.741 37.382 35.051 37.336 17. = 40.344 8.204 2.969 27.657 23.278 73.708 10.082 93.m 14.380 10.%9 25.m 33.m 27.4= 4.725 26.812 32. .198 32.632 M.16a 4.527 86.86a 48.340 13.133 ::E 21.S. .728 51.020 .475 20.848 M.872 1.M6 2.108 5. m 1 2 3 4 5 6 7 8 9 10 11 E 13 14 M 16 17 18 19 20 21 22 23 24 25 0.210 11.768 25.= 49.38. 1972.037 10.334 66.3= W.123 37.7o1 14.671 33.336 2a.= 30.335 49.325 3.812 6.881 77. .672 86.575 1.307 19.141 30.812 21.335 59.042 7.087 40.341 11.823 33.01 6.8U 9. .856 U. 17.238 20.338 19. 145 m.2m 27. . .151 16.337 20. = 46.853 22.334 89.041 14.267 74.319 32.167 2.329 E4 .385 6.307 85.488 11.435 31. .562 35.707 37.425 1.542 10. .339 14.9.= 52.620 54.508 34.351 5.26.035 7. 498 0.582 39. W 31.836 7.828 13.281 7.578 6.703 73..1.505 79.338 n. . F-8 .685 24.204 28.750 18.386 34.838 14.432 29.Om .773 4.366 3. .759 67.264 32. ~ athematical Function$. Percentage Points of the X2-Distribution. .026 22.843 23.05 0.336 28. . .336 53. D.C.719 37.= 14.779 9.358 1.706 4.339 U.816 16. SK 18.278 49.277 M.408 7. w? 0.275 18.797 48.000 33. page 984and 985.= 2.136 125.578 22.036 39.758 20.851 11.12. 017 13.169 10.065 . .117 16.684 15. .343 9.217 27.757 2a.896 71.088 2.= 36.337 24.766 79.345 13.10 2.m 45.82n 45.542 24.739 60.307 23.201 82. . .32a 64.863 48. . U.547 9.8.540 61.213 1.410 32.642 46.W 1.885 40.114 18. 473 17.267 35.524 12.332 10.940 4.845 35. W 36.939 34.172 36.087 15.256 51.409 34.337 23.382 23.141 E 28 29 30 40 w 60 70 80 w 100 $:E 27.338 18.652 38.= 8.50 . W2 52. .086 16. = 0.605 6.733 3.299 10.085 10.154 88.091 21.2E6 18.103 . m5 36.268 49.357 4. m 13.236 10.571 7.335 6a. = 48.s7 41.W7 322.W 5.239 1. 0.255 5.829 U?.773 55.1% 10.452 13.828 24.635 2.31Z f13.5a8 29. .805 83.711 34.633 8.711 1.578 1o7.528 32.882 63.278 21. 549 13.879 10.4SU 91.449 ::% 9.342 10.052 55. .672 9.240 U.389 n.6U 30.764 43. .337 . .107 4.M5 5.819 31.179 =.339 30.337 22.467 2n.515 22. Q8 17.87s z.071 5. .90 0.323 2.577 88. .833 3.064 1.039 27.00393 .053 3.688 29. .

M 2.86 1.36 3.s2 l.Q3 2.25 2.04 4.05 2.00 3. .73 2.m 3.46 4.26 3.99 3.& 1.59 5.32 4.71 1.86 2.40 2.23 2.79 2.42 2.97 1.07 2.51 8. F-9 .30 2.53 2.20 2.28 2.11 3.M 2.39 3.86 2.8!3 1.16 2.22 3.3 2. . .33 3.42 2.77 2.41 3.34 4.67 3.71 2.93 2.37 6.57 3.45 2.EM II1O-2-I415 5 Mar 93 Table F-6 VALUES F Distribution .18 2.55 6.+0 3.m 2.38 2.31 2. Washington. 20 .07 2.96 2.53 2.48 2.38 2.66 5. .91 2.32 5.59 2.73 1.01 2.34 2.02 1. .49 3.30 9.67 4.42 2.40 2.67 1.77 1. page 987.U 2.20 4.03 2.22 2.81 2.96 4.31 2.71 2.36 2.64 2.78 2.18 1. 15 245.85 2.70 2.m 2.41 8.% 3.07 3.01 1.7 19.74 2.s0 1.37 3.U 2.15 2.85 2.38 2.4 10.67 1.9 19.92 1.20 3.88 3. 1972.26 4.s 2.51 1. M 2.94 1.45 4.0 19. .71 6. .= 3.22 2.47 2.85 1.28 2.79 2. 2 lm.22 2.U 2.80 2. Percentage Points of the F-Distribution.06 2.74 1.85 2.00 w : 3 4 5 6 7 8 9 10 23 u 1.% 2.9.75 1.56 3.96 1.10 3.62 4.32 3. 1 161.53 1.34 3.44 3.W 4.32 2.42 3.30 3.59 5.77 2.10 8 23s.32 6.26 2.74 2.49 2.63 2.28 2.65 1.82 2. .ls 3.09 2.03 3.10 2.78 2.99 1.48 3.75 1.35 4. U.05 3.05) .40 2.51 1o.63 3.85 2.23 2. Q 2.84 OF THE F DISTR1BUTION Values for Upper Fra&ntit&nuamr 6 234.% 2.57 2.02 1.81 3.22 3.65 2.94 2.54 2.01 1.48 8.28 4.01 2.16 9.11 2.88 1.15 2.25 1.04 2.62 2.26 2.21 3.29 4.74 5.88 2.49 2. m 254.89 5.m 1.53 2.33 3. D.C.s 3.84 2.79 2.10 3.34 3.67 .36 3.30 4.= 1.78 1.59 2.28 2.23 3.01 1.28 4.48 3.60 248.53 2.01 1.23 3.62 2.45 8.20 3.62 2.00 3 .19 2.25 3.65 6.84 u 33.74 4.31 2.13 3.62 1.94 5.74 3.55 3.46 8.39 1.92 2.35 3.46 2.07 3.93 2.90 1.3.33 2.43 3.71 3.07 2.2s 2.0 19.39 3.43 8.96 1.32 2.06 3.65 1.2 3.75 5% (a= .W 4. .28 3.3 14 15 26 17 18 19 20 225.s 3.84 1.79 5.s 2.2s 3.62 5.46 .75 4.32 . .% 2.07 2.41 4.9s 1.66 1.73 3. .= 1.69 1. 2.76 1.60 2.49 3.05 4. .72 2.33 8.70 2.38 3.1 19.9 18.94 1.17 2. .23 2.80 2.57 5.95 1.55 2.24 4. Handb ook of Mathematical Functions.64 1.64 1.62 3.27 2.18 2.91 4.88 2.66 4. 30 250.80 2.34 2.51 2.07 2.26 6.21 4.93 1.57 2.55 1.08 3.30 2.81 3.70 5.27 2.2 19.24 3.75 4.% 1.00 4.46 2.54 2.21 .W 6.53 5.07 2.15 2.5 29.37 3. . ~of 4 224.37 5 230.19 2.76 2.2 19. .9 19.S.47 3.36 1.69 2.3 19.M 3.21 2.71 2.57 22 22 23 24 25 m 27 a B 30 40 60 320 From Table 26.25 2.55 2.60 2.61 5.26 5.92 3.53 2.94 3.79 1.= 1.60 2.50 3.57 2.45 2.12 2.66 3.87 3.48 2.36 4. National Bureau of Standards Applied Mathematical Series 55.84 1.44 3.85 ::.55 2.49 4.28 3.6 19.74 3.18 3.45 2.75 1.87 3.44 3.Q 1.69 3.84 1.63 3.81 2.12 4.01 2.39 5.63 1.03 2. . .25 2.31 2.04 1.03 2.s 2.47 2.00 9.69 4.3a 4.17 4.33 2.07 3.35 2.13 7.01 6.63 4.84 2.74 2.23 4.26 4.08 4.51 2.68 2.96 2.35 4. Government Printing Office.59 3.08 2.04 2.43 1.41 4.09 z.74 I.37 2.43 2.45 2.85 4.70 2.14 4.2.76 4.66 2.11 2.W 3.39 2.01 2.37 2. .25 9.53 4.51 3..= 4.10 2.07 2. .46 2.m 1. .07 3. 60 =.

.s32 .559 .370 5.537 3.76a 4.455 3.765 1.881 3.7s 1.669 1..W 1. 5.349 4.364 3.432 2.= 1.235 1.422 3.369 1.516 2. .655 .912 3.671 1.685 1. .372 2.848 1.755 3.422 2.* .607 1.662 1. .647 779.~8 1.853 .380 2.202 1.101 2.375 4.695 3.524 2. .= .3S6 5.580 3.719 .434 5.1o5 3.310 3.= l.382 2.382 2.978 2.339 1.m .675 24.548 .a66 1.m 3.m 2.291 9.W7 1.51m 4.m 2.385 1.932 2.m 3.6Q4 .562 .362 1.532 2.638 3. ml .543 .540 .257 .542 2.m 3.623 1.62n 3.230 3.042 5.170 2.ao 4.406 3. 20.= 2.333 1.05 0.629 2.694 6.266 .875 .536 Z. . This table is based on samples drawn from a normal distribution.432 11.= 1.675 1.874 3.2a5 2.574 2.248 2.435 4.225 2.225 3.5s 2.658 2.227 3.261 := .576 9.460 ADJUSTMENT .5a9 3.086 2.665 1.231 2.65s 3.177 2.296 1.341 1. m7 .769 .689 2.625 3. .M2 3.024 3.957 3.462 3. .765 1. 25.2n7 4.3s1 2.271 2.923 2.261 3.743 2.563 4.030 4.223 3.851 .496 3.1 0.259 2.278 6.43 8.388 3. .222 2.oa7 3. 38.451 1.965 1.980 1.3= 2.283 3.067 &.637 4.570 .542 . --.344 1.769 4. 9.309 .263 .433 4.268 .326 369.845 3. .ms 1.7M 1.681 1. .737 3.487 5.085 2.736 5.269 .3m 1.645 4.177 .347 1. 0.530 5. .446 4.452 3.633 1. 19.= 1.m .656 2.773 1.837 2.= .284 .394 1.069 2.807 1. .790 6.791 4.721 3.647 2.297 4.561 2.501 1.787 1.226 3.031 1.096 2.713 2.535 3.%30 4.775 1.333 .= 4.877 . .376 2.= 2. ml .256 .462 5.02 2 3 4 5 6 7 8 9 10 11 32 lz 14 u 26 1.090 7.149 5. .300 2. =3 &o .337 1.676 5.704 3.796 1.524 .644 1.178 3.11o 4.879 2.664 3.741 1.856 9.m 2.433 1.230 2.548 4.263 .270 .276 4.557 3.940 3. m 2.181 3.295 1.328 2.e86 4.855 2.014 4.221 3.645 4.170 3.531 .287 .304 3.8W 1.= 3.662 3.037 3.619 2.967 3.648 . .372 2.901 3.094 1.074 4.519 5.131 4.357 1.481 4.52a .492 6.202 3.276 .366 2.063 194.263 30.202 3.644 4.2M .W 2.= 2. .528 .265 .835 3.654 3.68U .549 .286 2.M 3.312 1.206 2.282 Note - Values have been generated by use of computer routines for the inverse t-distribution.551 2.782 2.340 10.m 1.354 3.651 1.279 .376 1.631 2.m 2.335 2.417 3. .044 4.660 2. 42n 14.5?2 6. .631 .160 3.544 .579 7.918 .m7 1.744 4.01 .113 2.-.EM 1I1O-2-1415 5 Mar 93 Table F-7 DEVIATES FOR THE EXPECTED -o 2.870 .324 4. .1.335 1.472 1.2.293 5.567 2.678 8.a84 2.= .161 4.010 1.297 100 m m m .189 4.6a2 2.m 1. . .539 1.535 3.234 4.862 1.078 3.793 3. W{ 11.303 1.396 .a32 2.507 4.751 1.618 4.7U 3. .964 3.734 1. The above values (after adjustment for skew) may be used as approximate adjustments to Pearson type-III distributions having small skew coefficients.873 . .833 2.529 . .250 4.641 2.819 1.341 4.575 .208 5.403 4.030 4.737 1. S46 .419 6.282 .823 3.957 14.707 2.326 4.772 1.670 3.720 3.= 1.= 2. .7= 1.W 2.328 2.700 3.611 3.074 3.%3 2.469 4. .761 /L0 OA.= 3.619 1.733 1.883 2.541 .999 2.850 .226 3.077 2.482 3.871 .464 2. . 10.354 3.779 1.” “.679 1.7= 2.2.264 3.7s 3.077 4.689 1.120 3.625 7. 0.463 4. . ~ N .927 .703 4.545 .742 2.887 2. 7.544 .304 5.W 1.260 .594 4. .1.822 2.258 1.790 1.M 3.261 .972 PROBABILITY ~o 0.257 .649 .061 3.979 3.397 3.051 3.487 31. 36.772 3.464 .%7 5.689 .748 3.035 2.229 2.552 .533 .901 .551 3.a54 2.607 17 la E 23 22 23 24 25 26 27 28 29 30 40 % w 70 80 90 .602 2.785 3.022 3.2423696.503 2. .842 1.8s5 1.408 3.367 Z.746 3.840 2.389 2.482 3.272 . .53s .078 2. .396 1.a49 0.W5 2.466 3.226 3.6% 1949.360 3.053 2.911 1.m .5 11.510 4.653 1.669 4.005 2.686 .950 .439 3.971 2.%2 6.775 3.126 2.554 77.906 .m 4.999 3.353 1.733 3.338 2.713 1.675 1.620 2.905 3.328 3.657 .5s2 .274 .= 4.680 1.2 16..066 3.649 l.745 1.874 .314 5.m 2.370 1. .730 1.391 4.499 3.034 7.285 2.* 4.529 .749 1. F-10 . .676 .381 .913 1.054 .7= 5.551 .781 1.727 1.449 19.173 3. ..126 5. .683 3.429 2.355 3.003 6. .567 5. .300 l.866 .5a2 2.067 3.256 .526 3.968 .864 ::Z 2.076 4.

63 2.63 7.42 .% 4. .7 M. 20.06 .m . 40.1 40.23 6.080 .oll 11.2 20.67 1.U .17 2.06J3 .28 1.33 3.19 3.9 2:.19 4.3 40.1 31.029 .88 5.7 40.54 1.33 6.080 .28 2.61 .3 40.88 .072 .4 M.0 FOR THE EXPECTED PROBABILITY ADJUSTMENT .04 5.5 40.3 40.1 40.2 30.311 .9 24.80 2.055 .13 1.61 .370 2.02 0.3 22.8 ~.453 .87 4.29 .5 30.43 17.41 8.31 2.016 .1 31.638 .050 .271 .o 11.219 .25 .232 . .01 20.1 40.188 .9 30. .m .73 1.6 30.217 .6 30. 37.s .32 7.X3 . 1.495 .7 20.3 40.2 10.58 .409 . .6 23.6 11.160 .018 2.11 1.5 14.947 .164 .45 6.85 1.81 lz.28a .130 .84 4.8 22.146 . 22.X 5.91 4.313 . .5 30.56 1.1 34.6 30.70 4.14 2.6 31.204 .8 30.62 . .255 .363 .09 3.05 0.076 .9 loos 10.145 ..598 4.531 2.710 . .228 .143 .47 .9 20.28 .73 5.8 31.127 .8 30.41 5.12 1.0 40.52 .39 4.26 .60 1.92 3.89 .37 2.3 10.3 21.089 oe3 .63 2.36 2.106 .77 5. F-11 .68 6.384 1.08 5.39 6.M 6.032 . .1 40.133 lzb .4 40.2 0.71 3.26 .91 .3 30.6 13.66 2.22 1. N 2 3 4 5 6 7 8 9 10 11 U lz 14 15 16 17 ls 1.09 3.097 .0 12.70 .1 30. 0.09 1.90 1.4 40. This table is basedon samples drawn from a normal distribution.36 .0 23.27 .97 1.57 6.4 20.3 40.7 ::: 33.019 .2 20.421 .599 .3 41.3 40.3 32.1 30.279 .427 1.52 1.02 1.47 2. .6 1.403 .164 .374 .244 .23 4.68 2.339 .049 .261 .197 .039 .82 5.X 2.= 2.23 3.45 .59 .7$3 5.993 1.2 31.00 4.3 19.08 11.9 30.0 30.1 11.26 4.77 8.0 40.00 8.5 11.18 l.lz 5.61 5.179 .I.4 41.4 11.03 2.20 1.69 2. ls 1.1 30.30 5.86 9.0 12.47 4.22 1.27 5.030 .7 30.116 .069 .87 5.65 .95 6.70 5.183 .20 4.050 .2 10.% 5.44 .23 2.073 .17 1.345 .12 =.56 7.22 4.32 .363 .24 5.= 2.68 10.7 20.36 3.m .1 24.4 1.36 4.1 40.* 2.389 1.= .2 40.o 10.o lz.6 m.1 0. .3 40. Ml .101 m9 1.7 30.1 40.EM 1110-2-1415 5 Mar 93 Table F-8 PERCENTAGES .1 40.020 .35 5. .2 30.450 .43 14.083 .4 22.56 .= 5.0 31.2 lz.053 .4 10.090 .110 .7 40.92 1. Ml .1 22.30 7. ox . 5.44 3.18 6.577 .43 .74 8.2 40.1 40.ss 5. 43.75 . .9 m. .= .3 20.061 .919 .085 .26 2. .2 41. 2< 30.Z 9.00 1.036 . .296 .086 .13 7.47 6.04 1.33 ::t S.6 23.168 .20 5.15 1.289 .336 .3 31.043 .020 .2 12.6 30.066 .3.102 .n 6.41 5.1 16.93 1.97 1.% 4. .5 42.9 m.1 22.74 2.230 .068 .32 6.4 40.173 .6 40.61 .33 4.3 2.01 1.9 40.2 20.66 .U 0.204 .49 .22 2.71 2.X 1.14 6.68 .3 14. . U9 .67 1.031 .E6 .1 11.077 .342 10.7 30.7 22.355 . .2 30.44 .m 1.85 5.249 .010 22.54 .4 40.602 10.02 1.5 40.25 6.65 .0 lz.24 3.a 20.100 .2 20.022 .23 2.052 .62 3.0 22.768 1.22 1.7 1o.0 a.2 30.W .49 5.54 .498 1.64 .258 .00 .M .3 10.184 .88 7.M . n 2.24 7.5 31.5 20.117 .8 41.2 22. The above values may be used as approximate adjustments to Pearson type-III distributions having small skew coefficients.84 2.81 4.3 32.05 10.50 Note - Values have been generated by use of computer routines for the inverse normal and inverse t-distributions.018 .141 . kmt-eb~e 4.0 1.3 11.658 .35 1.lz 1.396 .4 40.44 10.2s .22 5.242 2.00 5.16 4.48 1.64 1.50 .03 4.47 6.2 11.61 5. .259 .59 .11 3.38 9.3 20.9 a 21 22 23 24 25 26 27 28 2s 30 40 50 m 70 60 G 110 m m .036 .W .212 4.39 6.017 .75 12.W 3.3 u.219 4.59 .104 .4 20.78 4.4 21.80 1.5 40.1 23.5 30.099 .35 2.071 .485 .596 1.9 10.3 10.5 30. ls 2.496 3.6 40.0% .4 11.8 U.20 2.716 . 10.71 1.4 30.78 5.03 5.623 l.9 12.167 .51 5.IZ6 6.s 3.5 10.35 .95 1.526 .31 6.388 .08 6. .3 40.99 4.46 .52 33.330 .13 .516 .324 .816 .4 32.034 .5 11.37 19.Q 2.29 .78 1.98 2.

987 3.202 -1.690 2.577 3.388 4.056 .302 2.= -2.314 -.688 2. .062 -1.322 100 -1.639 4.O 2.160 142 0s -1.450 .317 .260 .535 3.I.667 -1.723 3.554 -1.774 ‘2.068 2.12s -2.602 2.= -2.m .785 4.4= 1.840 1.0 .244 -1.403 .064 1. 7a7 .347 3.fi9 -2. m? -1.250 -.107 2.229 -.371 2. .n4 3.865 ‘1.705 -.951 -1.309 -.706 -1.371 5.024 2.979 1.421 1.480 -1.460 -1.= -.= 3.787 1.405 1.%1 .9a6 2.708 1.ma .741 1.975 3.899 -.179 3.038 2.518 1.=9 2.53s 1. = -1. .719 3.Z5 2.970 1.= .946 2.250 1.644 2.079 2.229 3.344 1.951 4.744 -1.5 . -2.m .674 -1.552 .273 -1.254 1.641 1.265 -.9al 5.033 2.340 -.705 . .367 2.983 -1.= m -3.823 2.622 -1.350 -1.902 -1.604 .% .048 4.751 -.W2 3.894 5. Olz .220 m -2.359 .460 .51 2.212 3.282 1.= 1.321 -3.500 2.6Y3 -2. .666 2.491 3.62a mdalc -.m 2.1= 1.601 -2.4% -.588 2.990 25 30 40 xl 60 70 60 9) lW -5.556 1.EM 1110-2-1415 5 Mar 93 Table F-9 coNFIDENCE ~ N LIMIT DEVIATES FOR NORMAL D1STRIBUTION 99% and 98% Limits hmtkek~ .147 4.536 1.2 .764 2.535 .%2 4. 50 -1.202 1.955 -. -2.378 3.0“ i:o” -1.866 1.555 3. Zal 1. .632 -.032 -1.686 -1.379 60 -3.177 -.538 3.937 -.675 3.= -.464 1.320 2.135 1.236 1.657 1.768 2.656 4.561 2.446 3.266 -.103 -.131 2.175 .322 1.243 1.388 4.002 2.605 5.622 -1.743 1.457 1.769 : -4.= 2.336 1.637 1.442 3.326 = lo : . .639 ‘1.636 -2.= 1.038 -2.224 -.936 2.301 2. -1. .232 -.232 1.045 2. .975 5.603 3.757 -.49 -.35a .523 -1.= 2.248 2.274 1.063 2.611 1.682 -1.707 -.832 4.591 .szl 3.691 1.027 1.460 3.529 2.861 -1.3W 1.423 -1.m 1.563 -.Z31 3.030 -1.311 l.457 -3.161 4.419 .722 -1.279 -1.m l.581 -4.%5 .s9 2.212 80 -1.833 -1.866 .804 2. .109 1.232 1. .$84 .236 -1.= 1.67b 1.0 1.205 2.368 2.640 -3.508 4.048 -.009 1.W 2.26a 1.238 .806 2.252 -2.641 -3.974 2.235 2.384 2.= .996 .0 “ “b:o” “ “4.W 1.0 5.050 1.668 1.276 -1.346 -.249 3.236 .212 3.554 1.385 1. .517 5.503 40 -3.447 3.W8 1.172 -1.344 70 -3.=4 4.272 1.222 -3.355 -3.= 10 -1.651 1.005 e hl Calfidmc -.923 1.934 -2.U 3.028 -4.2n 100 -2.074 4.172 2.7= 1.255 4.530 4.690 1.867 2.364 -.561 6.372 3.552 .317 1.950 2.519 2.IZ6 6.s -.857 -.106 -1.504 5.706 4.851 2.m 30 -1.618 4.757 .3= 3. . .m .= -3.251 1.350 2.460 4.132 -1.224 1.340 1.966 -.2s4 -.367 ‘2.642 3.450 -.675 4.142 .403 -.102 2.202 -2.875 3.136 2.307 1. .207 .lc6 2.243 -4.m 1.364 .323 4.= -.346 1.999 3.3.624 2.645 1.314 1. 3a3 -2.223 3.832 3.226 2.030 2.273 2.785 3.062 1.412 .738 -3.a64 -.447 -2.632 -2.516 .= -1.059 -2.224 3.372 -2.651 .304 3.Zoa tiide .125 3.172 .no 3. .932 1.1o7 .m 1.0 10.42a 50 -3. -3.421 -.480 2.2s4 1.539 -.254 3.006 : 25 -1.458 1.633 1.788 -.423 3.048 -2.6s4 -.606 -2.974 1.H 1.= .970 . .0% 2.4a8 -.431 2.632 2. .823 -1.6M -1.178 5.245 70 -1.688 -1.258 2.11.550 -1.874 1.708 -.=6 -.249 -1.671 1.578 .5S 1.038 1.317 .1 : 20 25 30 40 30 60 70 80 90 100 -.710 4.038 -2.= -.382 2.= -.851 -.278 3.574 3.340 .565 .4% .= 1.606 -1.776 -1.530 -3.691 2. -.364 .005 3.630 1.010 e xl 3.6M 1.578 -.= 1.563 .= .= lcm 1.678 .336 4.741 -1.787 -1.132 2.245 1.4= .276 2.=7 1.863 Z. m 90 -1.313 -.%8 . .%7 1.423 2.* -.367 4.776 2.952 3. 2a6 -2.667 1.236 -1.751 1.283 -1.654 . m -1.651 -1.001 4.183 -2.224 -1.056 -1.738 -1.693 -. . .531 1.269 -1.285 -.738 Z.379 1.458 -.8a7 1.2n 2..317 60 -2.346 2.653 1.914 3.661 2.040 1.633 2.754 1.338 2.640 1.288 2.823 1.801 3.428 1.781 2.7= 3. .748 1.797 -.285 m -2.665 1.230 1.259 -1.lW 1.578 2.= -.048 3.636 3.738 2.329 3.722 2.460 .948 5.791 3.058 3.622 1.590 1.747 1.160 2.236 -.524 -.424 2.916 1.313 .= .lm -.364 1.172 -1.446 1.4a .524 .616 1.028 2.222 3.074 -3.336 ‘2.780 7.765 1.362 -2. .m 4.457 .567 3.55s 25 30 -3.764 -1.0 2.a35 l.535 .718 -1.in 1.823 2.789 1.21s 2.785 -2.678 -.625 -1.572 4.107 40 -1.09 4.731 -1.704 1.175 .53s -2. a74 -Z.364 -1.601 -.056 Lavel .5al 2.825 2.364 -.s33 -.107 -.259 .m -.8a3 -2.604 -1.= 10 -5.355 l.= 3.294 2.4U 2.554 6.325 .526 .56a -.272 w -1.426 .071 2.277 1. “ &:o” “ “95.=4 -.2b9 2.559 1.800 1.891 1.463 1.216 -1.647 3.744 4.654 -.678 -.4= 1.m -1.416 F-12 .650 -1.466 2.686 -1.263 :E .W -.916 -.545 -.483 -.436 20.365 -.229 -2.097 3.aao 1.551 2.143 6.708 2.744 1.718 3.829 -.125 -1.702 2.@ 3.2a5 1.225 -2.

759 1.040 -1.878 1.642 1.878 2.4% 2.797 -2.120 2.* -1. 10 M 20 25 .911 -2.5S 2.724 -1. * -.869 1.414 F-13 .373 .049 100 -2.791 4.286 1.0 “ “ &:o” “ “Go “ “ b:o” -..997 -1.801 1.657 1.738 2.284 -.376 3.999 2.022 3.035 1.W1 .040 1.374 .!370 -.733 -1.044 1.253 3.442 1.4* 1.146 ‘2.223 4.726 1.2K1 2..068 1.209 3.078 2.344 -1.9U -.348 i.765 2.565 .0s0 10 -1.= 1.298 .993 b.266 .436 -.864 3.m3 4.236 -.3W -.M -.483 4.801 -1.109 3.W6 1.227 -.571 3.831 .145 2.301 1..216 -.911 2.3= .s 2.777 -1. 7m -1.572 1.106 1.473 2.014 .351 -1.662 -3.396 2.73 2.317 M -1.407 1.151 2.510 3.:% 3.049 Lu’fOl - .310 2.696 -1.568 -2.014 -1.840 1.723 1.108 -2.079 1.320 ..053 2.:R ::R 4.220 .954 -1.883 3.896 1.124 -2.436 .652 1.7E6 3. W -1.449 1..= -1.482 -3.353 3.879 4.6e7 1. 5..950 .220 -1.s03 -1.733 2.353 -2.235 1.632 70 -1.233 3.= -1.99 1.= -*O m .158 -2.722 .667 4.468 -.064 2.355 2.W1 2.25s -2.= ..142 4.581 1.524 2.697 2.%7 2..026 -1.390 -1.652 2.917 1.679 -1.022 -.328 .1= -.786 .987 2.320 -.195 2.2s7 .373 -.091 -.326 1.694 1.3= -.1 .391 1.399 3.822 -1.= 2.m -.0 2. 7M 2.325 -1.095 -.831 1.646 -1.2 .068 -..226 -2.527 -1.442 2.344 1.757 1.017 -.5 3.865 3.468 .850 1.710 3.915 1.914 3.301 -3.266 -.672 5.s1 1.721 -1.596 1.539 2.077 -.711 1.931 -1..409 3.146 -2.601 3.375 2.2s7 -.380 1.677 1..387 -.342 -. 85s -2.576 2.647 2.= .629 .829 1.976 % -1.335 -1.= 2.911 2.284 .3s6 .976 .061 m -1.023 1.605 1.645 -1.762 1.373 .776 1.464 2.301 1.‘1.554 -.638 -3.980 2.838 1.005 2.702 -3.m 2.076 -2.455 -.077 1.390 .054 1. 819 -2.652 m -2.605 1.328 Fucmtaulcoh~e 10.134 .515 -.078 -2.295 -2.040 -.4= 2.632 1.* 1.559 -1.885 -1.= 1.182 1.549 3.403 .757 1.636 1.684 1.415 2.52s 1.396 -1.a33 1.252 2.916 .668 3.017 .921 -2.a6a 3.606 1.226 .590 -.263 1.224 -.980 1.975 -4.973 1.081 Wdmce .258 -.317 .EM 1110-2-1415 5 Mar 93 Table F-9 (Cent) co NFIDENC E LIMIT DE VIATF S FOR NORMAL DISTRIBUTION 95% and 90% Limits ~ U “ &:o” “ “9s.W 2.386 -1.= 1.222 .387 .113 2.482 1.= .320 .866 -1.440 3.= -2.695 3.761 3.712 -.313 4.145 1.563 -1.063 -1.le2 -..866 -1.628 3.m 2.174 1.310 -.423 3.572 -1.360 3.515 3.542 1.366 -.214 2.862 -2.065 -1.146 1.0 ~q 1.560 3.252 1.460 E -1..819 2.028 1.9E6 -1.354 .970 1.W7 2.454 2.= 1.480 .252 -2.964 1.759 -1.061 1.936 -1.772 3.468 .566 2.489 2.547 .898 -.62a 1.529 -2.822 -2.654 -1.607 4.414 -1.440 -3.859 2.= 2.840 -la -.442 -1.886 .618 1.807 -2.026 -.679 1.211 1.468 -.484 2.916 -1.342 .789 1.690 1.4W 30 -1.654 1.973 -1.8n2 1.628 2.370 1.:2 2.022 -1.831 -1.382 2.084 2.554 .232 -1.. m7 -1.39) 4..318 4.064 -2.526 2.370 -3.615 -1.874 3.s7 3.522 1.542 -1.423 -.569 1.232 2.809 2.520 -2.441 2.616 4.343 2.950 -1.351 2.647 -1.965 3.789 -1.578 Iavel - .654 3.0 .333 2. -.540 2.454 1.284 3.= 2.= .642 90 -2.404 -3.889 1.065 2.033 3.664 1.209 -.28s 2.5e2 60 -1.040 -1.034 2.542 2.040 ~1.786 -2.214 2.722 1.010 -1.286 .101 3.525 :: 1.032 -1.455 .478 3.823 1.486 1.175 -.944 1.733 -1..388 2.168 -2.745 1.162 2.145 -.854 1.684 -1.063 2.561 -1.228 . W1 -2.WO 2.161 2.096 3.W7 2.m -.250 .233 -2.026 2.542 5.357 1.m 2.933 1.576 -2.3s7 3.777 2..M 2.749 1.766 1.658 1.344 ‘3 .2n3 3.026 1.744 3.188 2.635 .028 -1.791 4.525 40 -1.103 2.405 -3.264 2.442 -.000 1.174 -2.578 -1.e22 3.618 1.227 -.328 3.267 5.60s -1. M -1.= -1.724 1.633 2.360 -1..m -1.W 2.360 1.= -.022 1.175 .715 .242 3.391 3.303 1.571 -.379 3.467 2.146 1.198 2.al -3.441 2.849 2.a7 4.073 3.421 2.076 1.039 3..365 50 -1.834 2.175 2.735 10 15 20 25 30 40 m 60 70 8n 90 100 10 33 20 25 30 40 xl 60 70 80 30 100 ‘1.673 2.= -1.= 3.2a3 -2.054 -.= 1.324 1.735 3.L% 2.= 2.264 2.235 2.232 -2.103 -1.336 2.= -1.580 .264 -1.0 1.079 -1.662 1.= 1.327 1.957 3.093 2.093 1.963 1.274 1.175 -.0s3 -2.126 1..006 -1.590 1.227 .403 -1.238 -.303 -.066 1.797 2.629 80 -2.856 -.297 .414 4.529 3.448 3.220 -.964 2.677 -1.354 -1.535 .808 3.955 2.605 -.238 1.718 2.223 -.989 2.m 3.433 1.032 2.258 .%7 -.757 -1.634 2.944 -1.765 -1.749 -1.158 3.4Q 3.425 2.433 .M5 1.175 .644 2.712 1.702 1.374 -1.304 4.=7 . 4m .618 -.618 -1.691 3.274 -.378 2.643 3.838 -1..344 1.011 2.= 4. ml -2..m8 -1.735 -..405 1.373 -1.400 2.9X .571 .786 2.786 -1.051 3.

363 3.S7 2.s22 1.0 1.m4 3.071 1.646 -1.926 1.010 2.479 2.366 -.262 1.270 2.* -.204 -.164 1.674 2.479 1.loa -.753 .m6 -1.4W -1.470 ~-e .065 -1.X 2.946 1.101 2. .071 -ma .919 2.433 3.o 10.262 -2.753 -2.627 2.206 -.671 -1.4 1. M.942 1.235 4.661 -2.3m .076 .425 -.793 2..638 -1.532 1.232 2. m -1.062 2.618 ‘1.061 -. .770 l.444 -1.772 2.331 3.632 2.336 2.154 1. m -2.762 2.463 2.202 1.43s 3.124 -1.666 2.482 3.474 2.240 -1.2n9 2.702 l.339 2.624 .W -. .75% .431 -1.414 2.S76 1. .0 .453 3.m6 1.320 1.625 -2.061 -1.mz .077 2. .1 3.320 -.077 -2.629 -2.206 -1.010 -2.775 2.144 2.634 -1.002 -2.m2 1.061 .764 -1.56a -2..W1 1.042 2.437 1.264 -.454 -1.4W 1.625 .263 2.O 3.u7 l.~~ek~ X1.093 1.633 2.44a -1.559 .475 -1.435 -3.L54 -.317 2.346 I.614 2.052 -2.0 5.596 .332 -1.155 -.266 -1.661 3.435 1.31.026 -.271 -.395 3.668 2.478 2.319 1.3= 1.084 2. .305 1.626 -1.322 -.667 -1.75o 1.206 2.679 3.724 2.731 2.3 2.7S 2.091 -2.5e3 -.429 .329 -1.W1 2.063 1.481 .940 1.m5 -1.942 .M5 2.347 1.276 1.426 2.601 1.662 .060 2. .66Q .795 -1.239 1.736 -2.720 .7s -1.536 1.100 2.354 -2.626 1.3m .601 -1.62.451 2.206 1.048 3.978 1.. =7 -2.m .561 2. 4a7 -1.063 n5 -2.221 3.476 -1.179 1.565 2.630 lam 1.614 -1.m 2.618 2.233 -.485 .104 -1.7W 2.600 2.906 1.m8 2.326 1.181 3.463 -1.306 1.445 .mh 1.357 2.702 -1.101 -1.273 2.018 -1.060 -1.6a8 2.mo 2.325 -.868 2.052 2.176 2.6s4 2.725 2. -2.= 1.7a2 -1.010 1.m2 -.362 -1. om -2. m ‘1.W 1.2211 3.347 -.062 3. -2.867 1.ml -.= .577 1.267 -.425 -1.362 -1.m9 -2.479 -1.639 2.662 2.066 Culfidme Level- 250 2. -2.667 1.978 2.173 l.251 -.46n 1.04 2.532 1. 75s .. us .812 -1.264 -.ml 1.767 1.201 1.544 1.077 -1.326 -1.160 2.740 2. Oal -1.194 2.772 -1 -2. -2.052 1.390 2.247 1.447 F-14 .466 -1.3m 1.220 2.511 -1.068 -.861 2.045 1.363 lam 1.740 1.009 -1.651 1.456 3. .765 -1.936 3.622 2.533 2.427 :27 l.333 3.200 .m .ml -2.756 2.323 3.s91 2.568 3.= 1.532 -2.606 -1.274 1.329 2.0 3.454 1.066 -1.523 2.* 1.173 -.s 2.=1 -1.ml 2.533 -1.336 1.324 3.761 2.475 .389 1.544 -1.332 1.767 -2.m4 2.631 1.250 1.657 2.222 -2.201 2.641 2. .276 -2.733 2.165 2. m -1.= 3. 46s -1.lm 1.541 3.179 -.793 2.369 1. . .m -1.M1 -.on 1.035 -2.223 1.511 .671 .513 3.674 -.320 3.810 3.819 2.855 3.276 3.474 -3.lo4 1.232 3.009 3.674 .570 -.2a4 2.627 -1.701 ekl= w~. o 20. .943 ~dmce kl.715 -1.147 2.763 m -2.s l.144 -.052 = -2.W 2.D 3.689 2.362 1.176 -1.261 -.747 -2.445 -.497 3.:E .357 3.m .m .434 2.470 -1.574 2.7s 1.538 -1.743 3.147 -1.532 3.441 2.612 2. m8 -1.701 lmve 1= .337 2.338 l.247 3.684 -2.715 lam 1.917 2.9U 2. m2 -1.702 2.657 -1.?m -2..661 .401 -. .624 -.076 -.538 -1.255 3.5m 3.425 2.731 4.235 2.351 1.2nl -1.603 1.735 2.120 -.103 1.794 3.874 -1.= -.577 -1.952 -2.203 &.437 -. -1.4a2 1.633 .765 1.517 2. .732 2.165 3.055 -1.s”””:2””’”.422 -.s33 2.363 2.261 3. u 99.106 2.0 2.432 .m 1.269 2.= -1.752 -.648 2.m5 -.708 2.3= -.ma 1.861 -1.043 1. .045 -2.= 2.222 2.326 l.522 -1.123 1.223 3.422 .668 2.2117 2.022 -2.431 1.623 2.m :Z .820 l.764 1.239 -.667 .747 .288 2.523 1.320 1.132 -1. m 2.629 10 25 20 25 30 40 3 60 70 60 m 100 10 M 20 25 30 40 50 60 70 30 20 100 10 M 20 25 30 40 w 60 70 60 20 lm -1.144 .167 1.470 -1.164 -.m l.018 2.662 -1.075 -1.126 -.267 -.132 -2.EM 1110-2-1415 5 Mar 93 Table F-9(Cont) c~ NFIDENCE LIMIT DEVIAT FOR N AL DISTRIBUTION 80% and 50% Limits fife ..944 -.4al -1.765 2.035 1.532 -1.517 2.055 2.763 .008 -1.122 1.342 2.144 l.470 “.036 1.278 -1.544 2.036 -2.= 1.002 1. .622 3.%4 2.085 2.m2 2.059 1. .946 -1.965 2.605 3.008 2.663 -2.290 -2.549 1.26-.532 -1.901 -1.657 .5e6 -1.321 2.235 -2.974 2.2Q9 2.085 3.4W -.735 -1.8a9 3.222 -.164 1.552 3.222 -.010 -1.=5 2.=7 2.739 2.501 2.am -1.720 -2.401 .7W 2.= -3.444 1.155 -1.517 -1. .m 1.634 2.645 -. .m5 2.664 .m5 1.223 2.075 1.539 3.364 3.058 3.207 -1.m -.o19 2.664 2.237 -.3.624 -1.647 4. .250 -2.3m 2.ea2 3.061 1.432 2.426 1.lal 1.271 .052 -2.002 -1.I.549 -1.070 3. m l(m = -2.873 1. .126 2.240 -2.479 1.940 2.240 1.986 -1.019 2. .541 :Z .. .355 1.043 -.222 1.225 1.425 :Z .042 -l+hu -1.0 95. .737 2.526 -1.m6 1.227 1.645 .422 -1.307 2.752 2.= 2.422 1.103 -1.m2 l.-1.736 .714 2.206 1.400 2.0 mo .105 3.231 3.627 2.146 -.

2a5 1.= 1.65E -1.064 Z.0 .651 -..357 -1.655 -1.6XI -1.846 -.563 2.m 1.076 2.262 -2. .692 2.365 2. .132 3.337 -1.666 -.645 -.910 2.289 1.m Midence .101 3.671 -1. .9a4 2.334 -1.o be W.645 -.062 2.6W 2.067 2.= 1.610 2.= 2.291 -2.100 -2.103 3.266 1.o C3umce*~e 10. .= 2.932 2.105 3.605 2.897 Z.645 .945 2.645 -.5 .652 -1.3S2 2.650 1.zm 1.662 -1.693 2.345 2.656 .163 3.266 -2.302 1.851 . B.635 2.335 -1. M6 . reference (25).324 -2. 1.286 -2.644 1.644 .m -2.m -.2 .266 1.846 -.619 2.m 2. b.650 . .060 2.656 1.702 1.647 .667 Source: Owen.070 2.854 .0 “ . .Ooo .341 2.111 3.593 2.651 1.357 2.345 -1.m .5%? 2.645 . .267 1.670 2.lW 2.265 -.0 . .66a -1- .927 2.0 “ b:o” ‘ “4.309 -2. .% 2.650 -.294 l.Om .654 -.335 2.652 1.301 -2.645 .6S 1.567 2. .Ow .116 3.EM 1110-2-1415 5 Mar 93 Table F-9 (Cent) co NFIDENCE LlMIT DE VIATE S FOR NORMAL DISTRIBUTION 50% Confidence Level size N 10 15 20 25 30 40 m 60 70 60 90 100 ‘ &:o” “ “9s.Ooo .0 .336 -1.379 2.267 -2.125 3.205 3.m .2s4 -2.379 -1.336 2.334 2.650 -1.m 2.107 3.646 .666 -1.365 -1.655 1.Ooo .410 -1.500 .061 2.m .301 I. D.661 1.S7 2.653 -1. F-15 .126 2.681 -1.352 -1.143 3.410 2.060 2.1. .917 2.339 2.339 -1.671 1.297 -2.Ooo .Ow .31 -1.647 -. 2.702 -1.0 5.066 2.1 3.= 1..Ooo .337 2.651 -1.

291 0.177 0.669 0.473 0.225 0.389 0.140 0.254 0. in Years (Nor H) .380 1. .127 0.358 0.257 0.243 1.383 0.532 0.313 0.157 0.131 0.365 0.768 0.254 0.504 0.080 0.066 0.067 0.5 0.231 0.4 1.253 0.148 0.225 0.128 0.657 0.417 o.126 0.351 0.167 0.479 1.550 0.175 0.505 0.113 0.1 2.994 1.073 0.912 0.120 0.314 0.127 O.303 0.552 0.183 0.754 0.764 90 0.666 0.0 0.977 1.064 0.990 60 0.7 2.0 1.311 0.178 0. . 7S2 0.6 0.442 0.313 0.167 0.118 0.667 0.326 0.109 0.272 0.116 0.243 0.221 0.1 1.924 0.3 0.315 0.261 0.928 0.403 0.866 0.240 0.0 2.84s 0.281 0.638 30 0.533 0.339 0. .376 0.292 0.689 0.862 0..481 0.432 0.114 0.338 0.192 0.EM 1110-2-1415 5 Mar 93 Table F-10 MEAN-SOUARE Station Skew (G) .066 1.704 0.160 0.836 0.212 0.103 0.082 0.473 0.089 2.851 0.425 0.181 0.610 0.634 0.123 0.117 0.562 0.087 0.201 0.624 0. .820 1.8 1.146 0.318 0. .422 0.1 0.532 0.412 0.623 0.244 0..880 0.716 100 0.496 0.174 0. .122 1.8 2.541 0.134 0. 10 0.176 0.485 0.093 0.529 1.142 1.808 0.119 50 0.167 0.282 0.624 0.186 0.143 0.239 2.4 2.701 0.821 0.139 0.399 20 0.690 0.272 0.654 0.295 0.0 Note . .4 0.542 0.975 1.077 0.3s5 0.7 1.316 0.315 0.397 0.376 0.202 0.410 0.468 0.479 0. 38S 0.204 0.252 0.534 0.488 0.047 1. .271 0.230 0.110 0.153 0.240 0.099 0. .580 0.513 0.095 0.442 0..259 0.196 0.498 0.347 0.513 0.621 0.074 0. 0s9 0.585 1.197 0.676 0.522 0.109 0.322 0.715 0.202 1.9 3.293 0.079 0.601 0.109 0. .692 0.910 0.412 0.275 0.2 1.185 0.166 0.083 1.068 0. . reference (46).489 0.252 0.4&7 0.410 0.086 0.9 2.741 0.463 0.751 0.262 0.5 1.419 0.739 0.071 0.080 0.570 0. .044 1.160 1.3 1.207 0.357 0.427 1.646 0. .523 0.223 1.101 0.135 0.494 0.010 1.572 0.058 0.283 0.093 0.6 1.561 0.106 0.943 1.613 0.063 0.6 2.456 0.290 0.716 0.340 0.543 0.543 0.211 0.086 0.2 2.603 0.582 0.059 0.713 0.105 0.728 0.823 80 0.075 0.093 0.156 0.150 0.lb? 0.158 0.292 0.332 1.440 0. F-16 .502 0. .794 0.069 0.054 0.Values computed from Equation 6 in Bulletin 17B. ERROR OF STATION SKEW CO EFFICIENT Record LenSth.552 0.5s1 0.224 0.507 0..9 1.187 0.137 0..200 0.359 0.5 2.449 0.631 0.288 1.285 0.2 0.089 0.130 0.8 0.359 0.507 0.334 0.465 0.311 40 0.391 0.363 0.766 0.335 0.644 0.092 0.780 0.163 0.589 0.679 0.102 0.383 0.211 0. ..142 0.100 0.471 0.235 0.&76 0..335 0.515 0.7 0.950 2.122 0.698 1.448 0.805 0.895 70 0.3 2.. .

385 2.981 2.109 3.591 2.775 2.961 2.138 3.116 3.114 3.126 3.931 2. reference (46).671 2.616 2.027 3.692 2.842 2.577 2.871 2.519 2.148 14 15 16 17 18 19 20 .089 3. Bulletin 17B.124 3.940 2.112 3.917 2.UFS (10 PERCE NT SIGNIFICANCE LEVEL) Sample Size 10 11 :: Outlier K Value 2.140 3.893 2.097 3.058 3.922 2.993 2.EM 1110-2-1415 5 Mar 93 Table F-11 Q~.061 Sample Size 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 Outlier K Value 3.052 3.736 2.175 2.067 3.335 2.700 2. 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 :: 67 68 69 70 71 72 73 74 75 76 77 78 79 Note - Table contains one sided 10% significance level deviates for the normal distribution.064 3.142 3.361 2.017 3.102 3.086 3.003 3.866 2.753 2..092 3.888 2.966 2.046 3.134 2.014 3.000 3..996 3.650 2.021 3.973 2.790 2.953 2.877 2.024 3.088 2.661 2. 23 24 25 26 27 28 29 30 31 :: 34 35 36 :: 39 40 41 42 43 44 :? 62 63 64 .710 2.563 2.055 3.036 2.081 3.279 2.070 3.100 3.849 2.534 2.006 3.897 2.429 2.719 Sample Size 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 Outlier K Value 2.818 2.448 2.146 3.798 2.213 2.903 2.860 2.549 2.037 3.760 2.049 3.970 2.073 3.949 2.030 3.133 3.135 3.122 3.040 3.768 2.033 3.783 2.078 3.628 2.467 2.682 2.011 3.727 2.945 2.935 Sample Siz e 80 81 82 83 84 85 :? 88 89 90 91 92 93 94 Outlier K Value 2.604 2.957 2.075 3.131 3.104 3.486 2.408 2.831 2.129 3.083 3. Source: Appendix 4.912 2.883 2.854 2.639 2.502 2.804 2. F-17 .144 3.989 2.984 2.247 2.043 3.977 2.107 3.744 2.309 2.824 2.811 2.119 3.927 2.837 2.095 3.908 2.

0333 .0000 .0006 2 3 4 5 6 7 8 9 10 .1771 .2199 .1076 .0036 . 40 100 50 60 80 . ..3585 . .0000 of Period in Years .0898 .0474 .1068 .1647 .0000 .1849 . . . . .1978 . .0 Number of Exceedances 0 1 Percent Chance Exceedance Event .0004 .1060 .0686 .0596 .0159 .0199 . .0490 .0044 .. .0264 .0649 . .1887 .0006 . .0148 . . Length 10 20 30 . 10.2059 .0052 .1285 .1055 .1148 .0643 .1471 .5987 .2277 .2702 .0000 .0812 .1319 .0019 .0105 .0596 .0120 .0000 .0004 .. .0587 .1068 .0020 . .1304 .0085 . .0339 ..0005 .0167 F-18 .0000 .1541 .3774 . ..0027 .0003 .0003 .1270 .1023 .0312 .0889 .0015 .0069 . .0000 . .0024 .0165 .0000 .2611 . .0133 .0001 .2146 .0246 .. . .0021 .0002 . .0119 .0000 . .0282 . .1937 . .1451 .1016 .0657 .0006 .2004 .0286 .1662 . .0527 .0574 . .2298 . .0319 .2586 .0027 .1032 .0001 .1851 .0180 .0010 .0016 .1455 .0000 .0779 .1451 .0022 .0104 .0105 .0000 . . .1423 .3151 .2003 .3874 .2706 .0112 .0000 .1603 . .1693 .0349 . .0001 .0000 .2361 .0058 .1500 .0461 .0576 . . . .0769 . .1809 .1216 .0746 .1800 . . .0001 .1781 . .3487 .0000 . .0001 .1396 .EM 1I1O-2-14I5 5 Mar 93 Table F-12 BINOMIAL R ISK TABLES Tables are probabilities of binomial risk for various exceedance frequencies and record lengths. .0342 .2852 .0000 .. .1724 . .0451 .0086 .1360 . ..3389 .0059 . . .1446 .2025 .. .0124 .0658 .1308 .0260 . .0001 .0889 . .1386 .1235 . 80 100 .0059 .1413 .0000 .0089 . . .0 Number of Ex ceedances 0 1 Percent Chance of Period 40 Exceedance in 50 Years 60 Event .. .1901 . .2259 .1336 .0424 .2777 .0844 .0000 ..0018 . .0000 .0901 .0393 .0695 .0152 .0389 5.0016 . Length 30 10 20 2 3 4 5 6 7 8 9 10 . .

.0000 .0000 .0000 .. .06 .0015 ..0000 .13 . .3638 ..0528 .0000 .00 .1986 3243 :2614 .1387 .01 .9044 .0002 .0001 .0000 .00 .0000 .0000 .0000 .0374 .0065 .0000 .0000 .8171 ..0000 .0000 .0159 .3716 .18 .0000 . .0000 . .0188 .0914 .1858 .0010 .0001 .0153 .0000 .1652 ..0000 .36 .0003 .03 .0000 . .0000 ..0000 . .0043 .0000 .6676 .0000 .0000 .. .4475 ..7397 .3316 .0000 of Period in Years .0028 .0006 .00 ~ N~ber of Ex ceedances 0 .0379 . Length 10 20 30 .0026 .6050 .3644 .0000 .00 .1448 .0000 .0000 .2976 .0004 ..0003 . .0000 . 0 Percent Number of Exceedances 0 1 2 3 4 5 6 7 8 9 10 Chance Exceedance Event .0545 . . . ..18 . .2242 .0002 ...0532 .0000 . 2.0027 .0001 . .0000 .0008 . .0006 .0000 ..EM 1110-2-]41: 5 Mar 91 Table F-12 (Cent) ~! TAB Tables are probabilities of binomial risk for various exceedance frequencies and record lengths.27 .00 .0865 .0042 .0011 .0145 .. .0000 ..0328 .0988 . . .0009 .0000 . ..0071 . .0031 . .0000 ..00 . .0000 .09 .0074 ..1667 . .0001 ..0000 .2194 .0000 .0988 .0000 .0000 ... .0000 ..0193 .0000 .6690 .0000 . .0000 .3340 .0000 .27 .0000 .1443 .8179 .00 .0000 . .0000 .0000 .00 0 1 2 3 4 5 6 7 8 9 10 .0000 ce E eedance Event of Period in Years .3056 ..0122 .0756 .0001 .0000 .0000 ..0000 .0000 . .00 . .0000 .5472 .0010 .0000 .0252 . 40 50 60 80 10 .. .3616 .0000 .01 .0000 .0000 . . Length 20 30 10 .0011 .0000 .4457 .0002 .3642 . 80 40 60 10 50 .36 . .00 .0000 . .0000 .2703 . .0058 . . .0000 .0607 .2725 . .5455 .0000 .0169 .0000 .0000 .0000 .0068 ..0000 .0000 F-19 .0000 .0000 .

0000 .0000 .0000 .8868 .7783 . .9046 .0000 .0000 .0000 .0000 .0000 .0000 . .0000 .0000 .0000 .0000 .0000 .0019 .0000 .0740 . .0000 .0000 2P ercent Number of Exceedances .0000 .0000 .0000 ..0000 .0241 . . . .1366 .0000 .0001 .0000 .0000 .6058 .0000 .0000 Ex ceedance in 50 Years 60 Event .0015 .1640 .0000 . . .0000 .0757 . . .0000 . .0032 . .2692 .0000 .0000 . .9608 .0002 .0907 .EM 1110-2-1415 5 Mar 93 Table F-12 (Cent) BINOMIAL RIS K TABJ.1645 .0161 .0000 .0000 .0000 F-20 .0000 Chance of Period 40 .0000 .0909 .0000 .0011 . .8520 .2232 .0000 .3044 .0006 . .0001 . . .0007 .0000 .0000 .0000 .9802 . .0001 . . .0163 .8183 .0000 .0045 .0000 . .0000 .0063 .0000 .0000 . Length 30 10 20 .6696 . . .0000 .0000 .0000 . . .0000 .7403 .8186 .0002 .0000 .0002 . .0001 .0000 .0000 .0000 .0000 .0000 . .. .0000 .0016 .0000 .0000 ceedance Event Number of ~xceedances 0 1 2 3 & 5 6 7 8 9 10 of Period in Years .0000 . .. . . .0000 .0000 .0000 .0000 .1956 .0000 .0566 .0070 .0000 .0000 .0000 .0000 .9417 ..0000 ..0000 . .0011 .0007 .0000 .0534 .0000 .0000 .0000 .0000 .0000 .1297 . .0000 .0001 .0029 . .0478 .0004 . . .0000 .0000 . 40 60 50 80 100 . .0000 .0000 . . .0000 .0000 0 1 2 3 4 5 6 7 8 9 10 . .0000 . .0000 . 80 100 . .0331 .9511 .0000 .0000 . .0000 .0000 . . .0000 .0000 . ~ ha ce .0000 .0108 .1066 .0010 .0124 .0000 . .0000 .0000 .0001 .0385 .0000 .0000 .9230 .0000 .0000 .0000 .0000 . .Es Tables are probabilities of binomial risk for various exceedance frequencies and record lengths. . .0000 . .0000 .0043 . .0001 . Length 10 30 20 .0000 . . . ..0095 .9047 . . .0196 .8604 .0000 . . ..0000 .0000 .0000 .

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd