You are on page 1of 8

# 3.

6
ESTIMATION OF MISSING DATA
Data for the period of missing rainfall data could be filled using estimation technique. The length
of period up to which the data could be filled is dependent on individual judgment. Generally,
rainfall for the missing period is estimated either by using the normal ratio method or the
distance power method.
3.6.1 Normal Ratio Method
In the normal ratio method, the rainfall PA at station A is estimated as a function of the normal
monthly or annual rainfall of the station under question and those of the neighboring stations for
the period of missing data at the station under question.
n

PA=

NR

NR A

i=1

* Pi

(3.11)

where Pi is the rainfall at surrounding stations, NRA is the normal monthly or seasonal rainfall at
station A, NRi is the normal monthly or seasonal rainfall at station i, and n is the number of
surrounding stations whose data are used for estimation.
Example 3.6: A catchment has four rain gauge stations A, B, C & D. Normal monthly rainfall at
these stations is known. The observed rainfall at the stations B, C & D for a storm event is
known and is given in the following table. Find the missing rainfall at station A.
Solution: The ratio of normal rainfall at station A to normal rainfall at station i or NRA/NRi has
been calculated and is given in table below.
Station

331.3

290.8

325.9

360.5

NRA/NRi

?
1

98.9
1.14

120.5
1.02

110.0
0.92

PA =

= 112.3mm
3

## 3.6.2 Distance power method

The rainfall at a station is estimated as a weighted average of the observed rainfall at the
neighboring stations. The weights are equal to the reciprocal of the distance or some power of
the reciprocal of the distance of the estimator stations from the estimated stations. Let Di be the
distance of the estimator station from the estimated station. If the weights are an inverse square
of distance, the estimated rainfall at station A is:

P /D
i

PA =

2
i

(3.12)

i=1
n

1/ D

2
i

i=1

Note that the weights go on reducing with distance and approach zero at large distances. A major
shortcoming of this method is that the orographic features and spatial distribution of the variables
are not considered. The extra information, if stations are close to each other, is not properly used.
The procedure for estimating the rainfall data by this technique is illustrated through an example.
If A, B, C, D are the location of stations discussed in the example of the normal ratio method, the
distance of each estimator station (B, C, and D) from station (A) whose data is to be estimated is
computed with the help of the coordinates using the formula:
Di2 = [ (x - xi)2 + (y - yi)2]

(3.13)

where x and y are the coordinates of the station whose data is estimated and xi and yi are the coordinates of stations whose data are used in estimation.
Example 3.7: Using the data of Example 3.6, estimate rainfall at station A using the distance
power method.
Solution: Since the coordinates of the stations are known, their distances from station A can be
calculated. The weights 1/Di2 are then computed for each station and the rainfall at station A is
estimated as follows:
Station
B
C
D
Total

Distance from
station A
28.0
17.7
42.5

1/Di2
1.29*10-3
3.19*10-3
0.55*10-3
5.01*10-3

Rainfall Pi
(mm)
98.9
120.5
110.0

Rainfall at station A =

Weighted rainfall
Pi*(1/Di2) (mm)
125.6*10-3
384.6*10-3
60.5*10-3
570.7*10-3

570.7 10 3
113.9 mm
5.0110 3

3.7
Disaggregation of Rainfall Data
Many hydrological applications, such as flood forecasting, require rainfall data of shorter
duration, whereas the network of recording rain gauges (providing short duration data) is small
in comparison to that of daily rain gauges. Hence, it is often necessary to disaggregate (or divide)
the daily rainfall data into shorter time intervals (usually one hour). The observed daily rainfall is
distributed in hourly values to follow the same pattern as the observed hourly rainfall. The

information of short interval rainfall is used together with the information of daily rainfall from
nearby non-recording (daily) gauges. A common method to do this is the mass curve method.
A mass curve is a graphical plot of accumulated rainfall at a station versus time. Mass
curves of accumulated rainfall at (non-recording) daily stations and recording stations can be
prepared by plotting the accumulated rainfall values against time for the storm duration under
analysis. Now the mass curves of the recording rain gauge stations are compared with those of
the non-recording stations to decide which recording rain gauge is representative of which of the
non-recording rain gauge for the purpose of distributing daily rainfall into hourly rainfall.
Assume that the daily rainfall is observed at 0800 hours. For converting the daily rainfall
into hourly rainfall, the hourly rainfall from 0800 hr to 0800 hr for consecutive days is
accumulated and the rainfall during each hour is expressed as a ratio of the total rainfall during
24 hours (0800 to 0800). These ratios for the identified station are used to distribute the daily
rainfall for the corresponding duration at the non-recording rain gauge stations.
The procedure for distribution of daily rainfall at non-recoding rain gauge stations into
hourly rainfall is explained with the help of an example.
Example 3.8: Daily rainfall data (mm) of four stations for the period 28th August to 30th August
1973 is given below:
Station name
28/08/1973
29/08/1973
30/08/1973
Pendra Road
65.3
80.8
1.1
Dindori
23.2
58.4
1.0
Ghansore
42.0
98.0
20.6
Niwas
30.4
203.0
33.4
Hourly rainfall data of two self-recording rain gauge (SRRG) stations (Mandla and Jabalpur) is
given in Table 3.5 for the period 27-29 August 1973. Disaggregate the daily rainfall data into
hourly values.

JABALPUR

Cumulative Rainfall
(mm)

300

200

100

MANDLA

27.8.73

28.8.73

29.8.73

NIWAS

Cumulative
Rainfall (mm)

300

GHANSOR

200

PENDRA ROAD

DINDORI
100

27.8.73

28.8.73

29.8.73

30.8.73

## Fig. 3.10 Mass curve of daily rainfall.

Solution: The hourly rainfall data at each of the two SRRG stations is plotted on a graph to
prepare the mass curve of hourly rainfall as shown in Figure 3.9. The daily rainfall data at each
of the four stations is also cumulated and plotted on a graph in Figure 3.10. The points are joined
to form the mass curve of daily rainfall. It may be seen that the mass curve of hourly rainfall has
many kinks as this has been prepared using hourly data. The mass curves of daily rainfall are
compared with those of hourly rainfall to determine which of the daily rain-gauge stations are
represented by which of the SRRG stations. Based on the match of curves, it was decided that
the daily rainfall stations at Pendra Road and Dindori are represented by Mandla while Jabalpur
represents Niwas and Ghansore stations.
To convert the daily rainfall into hourly rainfall, the hourly rainfall from 0900 hr on a day
to 0800 hr for consecutive days is cumulated and the rainfall during each hour is expressed as a
ratio of the total rainfall during 24 hours (0900 to 0800). These ratios are used to distribute the
daily rainfall for the corresponding duration at those rain gauge stations, which are represented
by the SRRG. The daily rainfall distributed for 1 day, i.e. corresponding to 28-29 August 1973 is
given in Table 3.6.
Remarks

It is very well recognized that the areal precipitation estimates derived on the basis of point
rainfall observations from a network of sparsely and unevenly distributed precipitation gauges
could only be regarded as an index of rainfall.
Radar sensed echo intensity reflected by precipitation could provide useful areal estimates of
precipitation over areas with few or no precipitation gauges. Precipitation observed by a dense

network of precipitation gauge specially set up for the purpose on a temporary basis could be
related to echo intensities as sensed by radar to develop reasonable relationships for the area of
interest.
Table 3.5: Hourly rainfall data of Mandla and Jabalpur
Hours
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

27/08/73
0.0
0.0
0.0
0.0
0.0
0.0
3.0
0.0
0.0
0.9
0.0
0.0
0.0
0.0
0.0
0.0
0.2
2.1
0.7
0.3
0.1
0.1
4.7
0.0

Mandla
28/08/73
0.0
0.0
0.0
0.0
0.0
0.0
1.7
9.0
2.3
0.0
1.5
5.5
0.0
0.0
0.0
2.5
1.5
0.8
3.4
0.6
2.5
0.0
0.0
0.0

29/08/73
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.5
0.4
0.1
0.0
0.0
0.0
0.0
0.3
0.4
0.3
0.0
0.0
0.0
0.0
0.0
0.0

27/08/73
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
25.7
0.0
0.1
1.0
6.0
1.3

Jabalpur
28/08/73
0.0
0.0
0.3
0.5
0.1
1.0
24.4
0.2
0.0
0.0
23.8
1.3
0.2
0.0
0.1
0.6
29.0
11.0
7.1
0.4
2.7
6.8
20.6
0.0

29/08/73
23.5
30.5
10.2
30.8
16.9
1.9
0.6
3.1
5.3
3.8
0.8
1.5
0.0
0.3
0.9
1.5
0.4
0.5
0.1
0.5
0.7
0.0
0.0
0.0

Table 3.6: Daily rainfall distributed into hourly rainfall (28-29 August 1973)
Hour

Pendra Road
28/8/1973 29/8/1973

0.0

Dindori
28/8/1973 29/8/1973

0.0

Niwas
Ghansore
28/8/1973 29/8/1973 28/8/1973 29/8/1973

21.5

10.5

2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24

0.0
0.0
0.0
0.0
0.0
0.0
0.0
8.9
0.0
5.6
21.8
0.0
0.0
0.0
9.7
6.5
2.5
13.7
2.4
9.7
0.0
0.0
0.0

0.0
0.0
0.0
0.0
0.0
0.0
0.0
6.4
0.0
4.1
13.8
0.0
0.0
0.0
7.0
4.7
1.8
9.9
1.7
7.0
0.0
0.0
0.0

28.0
9.3
28.2
13.4
1.7
0.5
2.8
0.0
0.0
21.9
1.0
0.2
0.0
0.1
0.5
26.6
10.1
6.5
0.4
2.4
6.1
18.9
0.0

13.7
4.5
13.7
7.5
0.8
0.2
1.3
0.0
0.0
10.7
0.5
0.1
0.0
0.0
0.2
13.3
4.9
3.1
0.2
1.1
2.9
9.1
0.0

3.8
ESTIMATION OF MEAN AREAL PRECIPITATION IN MOUNTAINOUS
AREAS
Precipitation data which exhibits appreciable spatial variation over relatively short distance is
often used as areal estimate for use as input in hydrological models. Several methods are
commonly used for estimating average precipitation over a specific area, such as a drainage
basin. The choice of the method is generally dependent on the quality and nature of data, the
importance of its use and required precision of the result.
Estimation of mean areal precipitation in mountainous areas by the conventional methods does
not yield the desired results because of the influence of orography, aspect, storm orientation, etc
on the precipitation in areas with pronounced orography. A better way of estimating areal rainfall
in mountainous regions is through the use of isopercental technique.
3.9

## STORAGE OF PRECIPITATION DATA IN DATA BANK

In India, precipitation data collected by central and state organizations are stored in hard copy as
well as on computers. All the rainfall data collected by India Meteorological Department (IMD)
are transferred to computers by the office of Additional Director General of Meteorology
(Research), IMD, Pune. Users can procure precipitation and other meteorological data from IMD
Pune (www.imdpune.gov.in). In addition, processed precipitation data are also available at some
internet sites.
3.10

## Rain Storm Analysis

While designing a dam, it is necessary that the outlet capacity is large enough to safely pass a
flood of certain magnitude. This critical flood is known as the design flood for the structure. The
type of the hydraulic structure is the main criterion to decide the design flood hydrograph. For
this purpose, the structures are classified as:
i)
ii)
iii)

## large or medium dams;

medium structures, such as barrages, road and railway bridges; and
small or minor structures, such as cross drainage works and minor irrigation tanks and
minor road bridges.

If long-term runoff data are not available, rainfall data which are generally available for a
longer period are used to estimate the design storm. This design storm is used with a suitable
rainfall-runoff model to obtain a design flood. The rainstorm analysis is the first step in the
design storm estimation procedure. The design storm (rainfall) is a magnitude of rainfall and its
distribution which is developed for the design of specific types of structures. It has three
components, namely, the rainfall amount, the areal distribution of rainfall, and the time
distribution of rainfall.
The Probable Maximum Precipitation (PMP) is defined as theoretically the greatest
depth of precipitation for a given duration that is physically possible over a given size storm area
at a particular geographical location at a certain time of year. The Standard Project Storm is
defined as that rain storm which is reasonably capable of occurring in the region of problem
basin. It is generally the most severe rain storm which has occurred in the region of the basin
during the period of available records.
The data on volumes of precipitation during severe storms is important for examining and
studying storms suitable for design purposes. Such information is generally presented in the
form of tables of the maximum average depth of storm precipitation over various standard area
sizes, such as 100 km2, 500 km2 etc. These data are known as the Depth-Area-Duration data and
they could be presented in tabular as well graphical form. WMO (2008) has described
procedures to estimate the probable maximum flood and standard project flood.

Closure
Undoubtedly, precipitation remains the most important and frequently used hydrometeorological in studies dealing with water resources. Precipitation is also the variable whose
longest measuring data series are available at a large number of stations. This module gives a
detailed coverage of acquisition and processing of precipitation data.

References
Jain, S.K., and Singh, V.P. (2003). Water Resources Systems Planning and Management.
Elsevier, Amsterdam.
Subramanya, K., Engineering Hydrology, Tata McGraw Hill Education Pvt. Ltd., New Delhi.
WMO(1982). Concepts and techniques in hydrological network design. Operational Hydrology
Report No.19, WMO No. 580, World Meteorological Organization, Geneva.
WMO(1982). Methods of correction for systematic error in point precipitation measurements for
operational use.
Operational Hydrology Report No. 21. WMO No. 589. World
Meteorological Organization, Geneva.
WMO(1983). Guide to Meteorological Instruments and Method of Observation. WMO No. 8.
World Meteorological Organization, Geneva.
WMO (2008). Guide to Hydrological Practices. Volume I: Hydrology From Measurement to
Hydrological Information. WMO No. 168. World Meteorological Organization, Geneva.