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

0

August 2007

www.gemcomsoftware.com

Copyright © 2007 Gemcom Software International Inc. (Gemcom). This software and documentation is proprietary to Gemcom and, except where expressly provided otherwise, does not form part of any contract. Changes may be made in products or services at any time without notice. Gemcom publishes this documentation for the sole use of Gemcom licensees. Without written permission you may not sell, reproduce, store in a retrieval system, or transmit any part of the documentation. For such permission, or to obtain extra copies please contact your local Gemcom office or visit www.gemcomsoftware.com.

While every precaution has been taken in the preparation of this manual, we assume no responsibility for errors or omissions. Neither is any liability assumed for damage resulting from the use of the information contained herein. Gemcom Software International Inc. Gemcom, the Gemcom logo, combinations thereof, and Whittle, Surpac, GEMS, Minex, Gemcom InSite and PCBC are trademarks of Gemcom Software International Inc. or its wholly-owned subsidiaries.

Contributors Rowdy Bristol Peter Esdale Phil Jackson Kiran Kumar Product Gemcom Surpac 6.0

Table of Contents

Introduction ................................................................................................................................................... 5 Requirements ........................................................................................................................................... 5 Objectives ................................................................................................................................................. 5 Workflow ................................................................................................................................................... 5 Required Files ............................................................................................................................................... 6 Tutorial profile ........................................................................................................................................... 7 Important Concepts....................................................................................................................................... 9 Understand the Domains .......................................................................................................................... 9 Check the Input Data ................................................................................................................................ 9 Understand the Estimation Method and Parameters ............................................................................. 10 Check the Output Model ......................................................................................................................... 10 Domains ...................................................................................................................................................... 11 A Simple Example .................................................................................................................................. 12 Viewing Domains in Surpac.................................................................................................................... 14 Extracting Data with a Domain in Surpac ............................................................................................... 16 Basic Statistics ............................................................................................................................................ 19 The Histogram ........................................................................................................................................ 20 Bimodal Distributions .............................................................................................................................. 22 Outliers ................................................................................................................................................... 23 Displaying Histograms in Surpac ........................................................................................................... 24 Removing Outliers in Surpac .................................................................................................................. 27 Anisotropy ................................................................................................................................................... 31 Isotropy vs. Anisotropy ........................................................................................................................... 32 Geostatistical Estimation Using Isotropy ................................................................................................ 34 Geostatistical Estimation Using Anisotropy ............................................................................................ 38 Ellipsoid Visualiser .................................................................................................................................. 43 Variograms .................................................................................................................................................. 53 Introduction to the Variogram ................................................................................................................. 54 Calculating a Variogram ......................................................................................................................... 56 Modifying the Lag Distance .................................................................................................................... 60 Omnidirectional Variograms ................................................................................................................... 63 Directional Variograms ........................................................................................................................... 64 Calculating an Omnidirectional Variogram in Surpac ............................................................................. 66 Modelling Variograms in Surpac............................................................................................................. 73 Variogram Maps.......................................................................................................................................... 85 Primary Variogram Map.......................................................................................................................... 86 Secondary Variogram Map ..................................................................................................................... 94 Anisotropy Ellipsoid Parameters ............................................................................................................ 96 Steps for Using Variogram Maps to Create Anisotropy Ellipsoid Parameters ..................................... 103 Inverse Distance Estimation ..................................................................................................................... 106 Isotropic vs Anisotropic Inverse Distance Estimation ........................................................................... 107 Steps to Performing Inverse Distance Estimation ................................................................................ 108 The Impact of Inverse Distance Power ................................................................................................ 113 Ordinary Kriging ........................................................................................................................................ 115

Page 3 of 137

........................................................ 124 Using Kriging Efficiency and Conditional Bias Slope ............................................................................................................................. 116 Impact of the Range .......................................... 134 Page 4 of 137 .............................................................................................................................................................................................................................................................................. 127 Model Validation ..................................................................... 133 Trend Analysis ................................................................. 131 Basic Statistics of Model Values...................... 128 Comparing Cross-sectional data with Model ............................ 129 Grade-Tonnage Curves .............................................................................................................................................................................Table of Contents Impact of the Nugget Effect .......................................................................... 123 Debug Output from Ordinary Kriging .......... 117 Block Size Analysis ............................................................................................................................................................................................................................................................................ 125 Block Site Selection ......................................................

Introduction Geostatistics is used in fields such as mining. 6. a professor of Statistics at the University of Wisconsin in the United States. “All models are wrong. you should have a good understanding of the following concepts in Surpac: 1. Geological Database Solid modelling Block modelling (how to create and constrain a model) Tcl scripts If you do not have a good background in these subjects. 5. the models you create are your responsibility. forestry. Also. Probably the most common use of geostatistics is to make estimations. A set of estimated points in space is known as a “model”. Objectives The primary objective of this tutorial is to help you become familiar with the methods for performing geostatistical operations with Surpac. Workflow The process described in this tutorial is outlined below: 1. and provide some guidance on making geostatistical decisions. Introduction Required Files Important geostatistical concepts Domains Basic statistics Anisotropy Variograms Variogram maps Inverse distance estimation Ordinary kriging Block size analysis Model validation Page 5 of 137 . 2. There are often more methods than those described here to obtain a model. 4. Ultimately. 8. 4. 3. once said. many parts of this tutorial may be difficult to follow. 12. This is often done in three-dimensional space. As George Box. 10.” Requirements Prior to proceeding with this tutorial.0 or later from a CD. hydrology. and meteorology in order to understand how data values change over distance. 3. 11. 9. Additionally. this tutorial will introduce you to some general geostatistical concepts. Some are useful. 7. 2. you will need to have installed Surpac 6. such as the specific gravity of rock for an area where there are only a few known sample values.

If you have not installed the software from a CD: 1.com/Tutorials. Unzip the file geostatistics. Page 6 of 137 .zip into the directory you created. Download the geostatistics tutorial/data (contained in a single zip file) from: http://www. you should have installed Surpac 6.asp 3.surpac. The files and directory structure for each tutorial will only be present if you have installed the software from the CD.0 or later from a CD.Required Files Workflow Required Files Overview This chapter will identify where you will find the files required for this tutorial. Requirements Prior to performing the exercises in this chapter. Create the following directory: c:/surpacminex/surpac_60/demo_data/tutorials/geostatistics 2.

3. listing all available tutorials. From this popup menu.Required Files Tutorial profile Tutorial profile Profiles are a collection of menubars and toolbars. A new menubar will be displayed. The tutorial profile contains a set of menus that assist you with learning various aspects of the software. Page 7 of 137 . choose Profiles > tutorials. Choose Geostatistics > CD to geostatistics folder. To display the tutorials profile: 1. 2. Right click in the blank area to the right of the menus.

Refer to the Introduction to Surpac manual for more information on profiles. Page 8 of 137 . Summary The files you will need for the remainder of the tutorial should now be present in your work directory.Required Files Tutorial profile This should set your working directory to: C:/surpacminex/surpac_60/demo_data/tutorials/geostatistics This directory contains all of the files required to perform the steps in this tutorial.

Assuming that the quality of the data is as good as you’re going to get. These concepts can be divided into four regions: 1. it is important to group all sample data contained within each domain into distinct subsets. Page 9 of 137 . and use data from each separate domain to make estimations within that domain. there are a couple of potentially hazardous characteristics of the data which you should look for: “bimodalism” and “outliers”. After that. you can analyse each subset individually. it is likely to contain more estimation errors than a model created from a unimodal data set. know what a geostatistical model is. Once you have identified the domains. You can look for both of these features with a histogram. Additionally. can cause estimation errors. the data is said to be “bimodal”. these subjects are outside the scope of this tutorial. If you use some of the more common estimation techniques to create a model based on a bimodal distribution.Important Concepts Understand the Domains Important Concepts Overview Although geostatistics is not an exact science. “outliers”. Understand the Domains It is important to recognise separate “regions” or “domains” within a model. 4. or values which are significantly distant from the majority of the data. Although sampling theory and laboratory quality control practices are important concepts which impact the quality of any estimation made using a set of data values. If there are two peaks. A data set is said to be “unimodal” if the histogram shows a single peak. there are some important concepts which can reduce estimation errors. 2. or have previously performed a geostatistical estimation. Domains Validation of input data Understanding estimation methods and parameters Validating the output model Requirements There are no requirements for reading this chapter. 3. but you may find some of the principles easier to understand if you: • • • have some understanding of basic statistics. Check the Input Data The saying “Garbage in = Garbage out” is certainly true in geostatistics.

you should have a good background in basic statistics. it is important that the pilot understand the theory of aerodynamics to understand what impact a particular control has upon the end result. as well as basic geostatistical principles. Page 10 of 137 . It is important to recognise that validation of input and output data are as important as understanding geostatistical theory and the estimation method being used. and a large number of parameters within each method. Although there are “autopilot” modes. where you just press a few buttons and something happens. cross sections of block models. contours of plans. colour coded and rotated in three-dimensional space are all methods which can be used to verify the output values.Important Concepts Understand the Estimation Method and Parameters Understand the Estimation Method and Parameters There are a large number of estimation methods. Check the Output Model A final method you should use to check the quality of estimation is to take time to examine the output. It is an inexact science which is used to make estimations at locations where no data exists. Before using a particular estimation method. Using geostatistics can be likened to flying a jet plane. Histograms of estimated values. Summary Geostatistics is the study of how data varies in space.

know how to display string and DTM files. or not classifying data into correct domains. Page 11 of 137 . Mixing data from more than one domain.Domains Check the Output Model Domains Overview One of the most important aspects of geostatistics is to ensure that any data set is correctly classified into a set of homogenous “domains”. 2. you should: • • understand what Surpac string and DTM files are. The following concepts will be presented in this chapter: 1. Estimation without domains The impact of domains on estimated values Requirements Prior to proceeding with this chapter. can often be the source of estimation errors. A domain is either a 2D or 3D region within which all data is related.

Page 12 of 137 . and C. and you are given three air temperatures measured at locations A. 2. Given the previous two points. 3. it is difficult to say with certainty if the temperature at X is above 25 degrees. and the estimate is exactly 25 degrees. as displayed below. they will probably have equal influence on the estimated temperature. B. Since locations B and C are about the same distance from X. you may have the following thoughts: 1. the temperature at X would probably be the average of the temperatures at B and C: (18 + 32) / 2 = 25 degrees Since the influence of A has not been accounted for at all. Based on the values shown. the value at A may have little or no influence on the estimated temperature at X. Since location A is relatively distant from X. what would you guess the temperature is at location X? Would you guess that the temperature at location X was greater than 25? What is the temperature at location X? Using the information above. 4.Domains A Simple Example A Simple Example Imagine that you are a meteorologist.

but only if the temperature is 25 degrees or more. because temperatures high up in the mountains are usually not good estimates of temperatures on the beach. You have three friends who live near the beach you want to go to. you may have the following thoughts: 1. The point is that separating data into similar regions. 2. you came up with a considerably different result.Domains A Simple Example Now consider the following: Imagine that you want to go to your favourite beach. and C) and the temperatures they give you. Your favourite beach is at location X. in the second case. with the locations of each friend (A. The data from B can be ignored. you would probably decide “Yes. the temperature at X will probably be somewhere between the temperature at A and the temperature at C. Page 13 of 137 . and you call them up and ask each one what the temperature is at each of their homes. Since X is between A and C on the map. A and C are on the beach. or domains is a very important part of making any geostatistical estimation. Therefore. all of the locations and temperatures are exactly the same. B. 4. I’m going to the beach!” Compare this example with the first one. Since the temperature range of 28 to 32 degrees is greater than the minimum value of 25 degrees. You draw the map below. when you took account of the domain which the data is contained within. so they can be used to guess the temperature at X. the temperature at X will be somewhere between 28 and 32 degrees 5. Would you go to the beach? Using the information above. Note that the friend at location B lives high up in the mountains. 3. However. while friends at A and C live near the beach. In both cases.

3. and then click Apply. 2. Choose Display > 3D grid. 5. and then click Apply. Enter the information as shown. Choose Display > Hide everything.Domains Viewing Domains in Surpac Viewing Domains in Surpac 1.str. 4. 6. Choose Display > Point > Markers. Enter the information as shown. Open all_composites2. Page 14 of 137 .

The D1 values have been used to classify the points into different strings: String 1 2 3 4 5 6 7 D1 < 1.999 4 – 4. any estimation that you would make with only this file would be based only on the distances between the sample points and the estimated location.str The points in this string file represent 2 metre downhole composites. Page 15 of 137 . all_composites2.000 1 – 1.999 >= 6.Domains Viewing Domains in Surpac You will see an image as shown.999 2 – 2.999 5 – 5.000 As in the first example above. The D1 field contains the composited value for gold.999 3 –3.

After reading the text below on the first form. 2. click Apply. you will be creating 2 metre downhole composites within the QV1 geological code. 1. as interpreted by a geologist. Extracting Data with a Domain in Surpac The domain ore_solid1.dtm This solid represents a single domain.tcl. With all_composites2. You will now go through the process of extracting composites only inside the QV1 domain. Run the macro 01_create_downhole_composites.dtm.str still displayed on the screen. In this example.Domains Extracting Data with a Domain in Surpac 7. Only composites which fall inside this domain should be used to estimate points inside the domain. ore_solid1. open ore_solid1.dtm represents an ore zone known as the QV1 zone. A geostatistical analysis of data in a drillhole database generally starts with compositing a sample value within a given geological zone. Page 16 of 137 .

dtm. Note that a composite length of 2 metres has been selected. After viewing the form below. After viewing the form. but is beyond the scope of this tutorial. notice that the character field rock has been set up in the geology table. You may want to consider the opinion of a geostatistical consultant to determine the optimal composite length for your data set. which is an interval table. Page 17 of 137 .Domains Extracting Data with a Domain in Surpac The function COMPOSITE DOWNHOLE is invoked using Database > Composite > Downhole. 3. The selection of a composite length is important. click Apply. 4. click Apply. On the next form. The text “QV1” has been inserted into the field rock for every interval of a drillhole which is inside ore_solid1.

2metre composites inside QV1 zone Summary You should now understand the impact which domains have upon geostatistical estimations. After reading the text on the next form. and how to use Surpac to extract data within a domain. click Apply. You will see an east-west section of the database and the composites which were created.Domains Extracting Data with a Domain in Surpac 5. Either or both of these strings may be used for further geostatistical analysis. Page 18 of 137 .str. String 2 contains composites where less than 50% of the 2m length contained a gold value. 2 metre downhole composites have been created within the QV1 rock type. and are stored in the D1 field in gold_comp2. String 1 contains composites where 50% to 100% of the 2m length contained a gold value. you will use both strings. In this example.

Using a histogram to identify a bimodal distribution. The following concepts will be presented in this chapter: 1. Selection of a cutoff value.Basic Statistics Extracting Data with a Domain in Surpac Basic Statistics Overview One of the important preliminary steps in performing a geostatistical evaluation is to have a good understanding of the raw data. A histogram can be used to identify both of these. Two characteristics which can potentially reduce the quality of your estimations are bimodalism and outliers. 2. 3. Using a histogram to identify outliers. you should: • • be familiar with Surpac string files know how to run a Surpac macro Page 19 of 137 . Requirements Prior to proceeding with this chapter.

0 – 4.0 2.5 6.1.0 4.5 2.0 5. a distribution of gold grades could be represented by the following table: Gold (g/t) Number of samples (frequency) 0.0 .0 .2.5 – 2.5 – 1.5 3.0 . value.5 – 4.0 – 7.5 7.0 3.0 6.5 – 8.3.0.5 – 3.5 – 6.5 1.5 – 5. A histogram is the graphical version of a table which shows what proportion of cases fall into each of several nonoverlapping intervals of some variable.0 1.5 4.5 – 7.5 0.0 7.0 – 6.0 . For example.0 0 40 58 82 40 29 18 10 12 5 5 5 5 8 5 Page 20 of 137 .Basic Statistics The Histogram The Histogram A histogram is a statistical term which refers to a graph of frequency vs.

Basic Statistics The Histogram This same data can be displayed in a histogram as shown: Histogram of gold grades Page 21 of 137 .

Page 22 of 137 . and the distribution is said to be “bimodal”: 1 2 2 2 3 5 5 8 8 8 9 Imagine that you are studying the average specific gravity. For example. In order to minimise estimation errors. merely segregating the data based on rock type would result in two separate normal distributions. in the following data set. is said to be “bimodal”. as in the example above. the numbers 2 and 8 are equally common. The specific gravity values between 1 and 2 are representative of the coal. the number 8 is the mode: 1 3 5 5 8 8 8 9 “Bimodal” means that there are two relatively “most common” values which are not adjacent to one another. A histogram of all rock samples might look like this: Specific Gravity Any histogram which displays two humps. In the example above.Basic Statistics Bimodal Distributions Bimodal Distributions The “mode” is the most commonly occurring value in a data set. In the following data set. or density of rocks in a coal deposit. The bimodal distribution in the example above can be explained by the fact that the data set is comprised of coal samples as well as intervening sandstone and mudstone bands. Often the source of a bimodal distribution can be two domains being mixed into a single data set. you should make every attempt to separate any data set which has a bimodal distribution. while specific gravity values between 2 and 3 represent the intervening rock.

Additionally. In the example above. For example. Page 23 of 137 .Basic Statistics Outliers Outliers An “outlier” is a statistical term for a data value which is relatively distant from the majority of all other values in the data set. One technique used to reduce the impact of outliers is to apply a “cutoff” to them. outliers can result in unrealistic results. if used in an estimation. the value of 236 could be “cut”. in the following data set. the number 236 would be considered to be an outlier: 1 3 5 5 8 8 8 236 Outliers can cause problems with the calculation of variograms. or changed to a value of 9: 1 3 5 5 8 8 8 9 Another alternative is to remove the outlier value(s).

tcl. Basic Statistics on gold_comp2. 2. the data should be analysed graphically to see if it can be physically segregated into two separate zones.Basic Statistics Displaying Histograms in Surpac Displaying Histograms in Surpac 1. each zone should be modelled separately. The value used to cut or remove outliers can be calculated from information in the basic statistics report. 2. click Apply.str Page 24 of 137 . If the histogram shows a bimodal distribution. File > Load data from string files is selected. If so. The quality of experimental variograms and subsequent block model estimations are sensitive to outliers (relatively large values). Outlier values should be cut or removed prior to variogram modelling or block model estimation. Run the macro 02_basic_statistics. Next. Basic statistics should be performed before variogram modelling for a couple of reasons: 1. The shape of the histogram can be used to determine if a distribution is bimodal (has two humps). The Basic Statistics window is opened by selecting Geostatistics > Basic statistics. and the form below is displayed. After reading the text below on the first form.

click Apply. click Apply. a histogram and a line representing the cumulative frequency is displayed. or Z coordinate values. click Apply. This form prompts you to enter the name of an output report. you can exclude data which is greater or less than any Y. After reviewing the form.str as the basis of our study. The basic statistics report will be displayed next. this distribution is not bimodal. Note the values of the mean. and percentiles. note that it is possible to view the histogram based on a number of bins or on a bin width. 4. The “bin width” method is more commonly used. The “Name” field is optional. Next. X. The columns labelled “Minimum value” and “Maximum value” allow you to exclude data which is below a given minimum value or above a given maximum value. The name value will appear on the output report. Report was selected from the Statistics menu. standard deviation. and a range of percentiles which will be written to the report. After reading the text displayed on the next form. On the Advanced tab. As you can see from the histogram. Basic statistics histogram and report 5. Also. After this. When you have completed viewing the form. the report format.Basic Statistics Displaying Histograms in Surpac You will use strings 1 and 2 from the file gold_comp2. The cumulative frequency is an accumulation of the values of all previous histogram bins. The D1 field contains values of gold in grams per tonne. Page 25 of 137 . 3.

832 1.not is displayed.0 Percentile (median) 75.120 9. This report contains several output statistics.Basic Statistics Displaying Histograms in Surpac The output report raw_gold.str ---------------------------------------String range Variable 1.867 41.672 6.0 Percentile 50. You will refer to this report in the next section.120 3. including the specified percentiles.483 Trimean Biweight MAD Alpha Sichel-t 2. Output Filename: raw_gold Statistics Report File Gold Comp2.658 2.2 Gold Number of samples Minimum value Maximum value 335 0.784 Skewness Kurtosis 5.490 25.0 Percentile 99.0 Percentile 90.280 44.728 5.299 2.94e+010 Page 26 of 137 .705 -0.0 Percentile 1.298 5.113 Mean Variance Standard Deviation Coefficient of variation 3.0 Percentile 95.730 63.828 46.235 0.

Variograms and subsequent block model estimations are sensitive to outliers (relatively large values). Run the macro 03_cut_outliers.96 * standard deviation) For this data set. 95th.120 9. One method of dealing with these data are to reduce.280 99th Percentile: 44. Recall the following percentile values were given in the basic statistics report: 90th Percentile: 95th Percentile: 5. 1. or 'cut' them to some lesser value. you will use the value which defines the upper limit of a 95% confidence interval A confidence interval is an estimated range of values which is likely to include a given percentage of the data values. Any given percentile value could be selected as the outlier cutoff. The value used to cut outliers can be determined by one of several methods. or an arbitrarily chosen value. After reading the text below on the first form.831) 95% CI = 17.Basic Statistics Removing Outliers in Surpac Removing Outliers in Surpac Looking back to the histogram of gold_comp2. such as the 90th.tcl.217 For simplicity. such as a percentile. The calculation for the upper limit of a 95% confidence interval (CI) is: 95% CI = mean + (1.828 and standard deviation = 6. A given percentile 3. Also. The upper limit of a given confidence interval 2. Since a confidence interval is based on the data alone.828 + (1. it is useful where there is little or no knowledge of the deposit.831 95% CI = 3.96 * 6. including: 1. as well as the output report. you can see that the majority of the data is grouped between values of 0 and 10 grams per tonne. 2. you can see that there are several outlier values above 10 grams per tonne. click Apply. you will use the nearest integer value of 17 to cut the outlier data.str. or 99th percentile. A percentile is that data value at which a given percentage of all other data values fall below. As stated above. An arbitrarily chosen value In this example. other methods can be used to select the outlier cutoff.112 Page 27 of 137 . mean = 3.

Whatever method is chosen. the macro has opened gold_comp2. you will analyse the data in the Basic Statistics window. STR MATHS is invoked by selecting File tools > String maths. information from grade control samples and reconciliation studies may provide a good idea of what the maximum mined block value will be. then set the value of d1 equal to 17.d1) This expression can be reworded as: If the initial value of d1 is greater than 17. Prior to viewing this form.17. When you have completed viewing the form. Using string maths to cut outliers In order to validate the output from STR MATHS.Basic Statistics Removing Outliers in Surpac An arbitrarily chosen value based on knowledge of the deposit and sampling methods may also be used. If the deposit has not yet been mined. This form prompts you to enter the name of the input and output files.str. Again. 3. this is invoked by selecting Geostatistics > Basic statistics. For example. information from similar deposits may be useful in determining the outlier cutoff.str and saved it as gold_cut17. click Apply. values in a description field in a string file can be cut with the use of STR MATHS. Page 28 of 137 . The D1 field will receive the result of the expression: iif(d1>17. if part of an ore zone has been mined. as well as an expression. else leave the value of d1 as it was initially.

Statistics > Report was selected. and a range of percentiles which will be written to the report. When you have completed viewing the form. click Apply. 5. This form prompts you to enter the name of an output report. Basic Statistics on the “cut” data set Next. Notice that gold_cut17. click Apply. the macro will choose File > Load data from string files.str is the file being analysed. a histogram and a line representing the cumulative frequency is displayed. the report format.Basic Statistics Removing Outliers in Surpac Next. After this. Notice that the maximum data value is now 17. Percentile range definition Page 29 of 137 . and the form below is displayed. When you have completed viewing the form. 4.

str contains the D1 values from gold_comp2. After reading the text below on the next form.120 3.str will now be used for all subsequent variography analysis.985 Skewness Kurtosis 3.000.str ---------------------------------------String range Variable 1.235 0.487 Trimean Biweight MAD Alpha Sichel-t 2. As displayed by this histogram. This file is created in the directory.298 Mean Variance Standard Deviation Coefficient of variation 3.str.not contains several output statistics.133 0. you can see that the maximum value is 17. but not displayed by the macro. and also how to select and implement an outlier cutoff. Page 30 of 137 . as well as block model estimation.0 Percentile 1. Output Filename: gold_cut17 Statistics Report File Gold Cut17.728 2996. click Apply. The D1 field in the file gold_cut17.728 Summary You should now understand how basic statistics can be used to identify bimodal distributions and outliers.814 3.Basic Statistics Removing Outliers in Surpac 6.200 13.730 17. including the specified percentiles. The output report gold_cut17.705 -0.2 Gold Number of samples Minimum value Maximum value 335 0.658 2.0 Percentile (median) 75. The D1 field in gold_cut17.0 Percentile 50.182 9. You may open it if you wish and verify that the maximum value is 17.000 25.299 2.

you should: • • • understand Surpac string files. Geostatistical estimation using isotropy. Requirements Prior to proceeding with this chapter. 2. Ellipsoid visualiser. and is described in this chapter through the following: 1. 4.Anisotropy Removing Outliers in Surpac Anisotropy Overview An important aspect of performing any geostatistical evaluation is to understand how data values change with regard to direction. The term “anisotropy” deals with this concept. anisotropy. 3. Isotropy vs. and how to display them be familiar with the geometric shape and deposition of economic geological deposits understand the concept of a centroid of an individual block in a block model Page 31 of 137 . Geostatistical estimation using anisotropy.

Anisotropy Isotropy vs. No Anisotropy (Isotropic) 2:1 Anisotropy Azimuth 45 2:1 Anisotropy Azimuth 135 5:1 Anisotropy Azimuth 135 Page 32 of 137 . having a different value when measured in different directions When estimating values in a block model. Anisotropy Isotropy vs. Anisotropy In order to understand anisotropy. For example. the three models shown below were created from the same data set. the amount and direction of anisotropy can have a significant impact on the end result. but different amounts of anisotropy were used. Here is a definition of each: Isotropy: the property of being isotropic. having the same value when measured in different directions Anisotropy: the property of being anisotropic. it is helpful to know what the term isotropy refers to.

Anisotropy 1. and perpendicular to the dip plane. an isotropic condition in two dimensions is more common. For example. down dip. Page 33 of 137 . Conversely. If you use the Macro playback button. For example. Run the macro anisotropy. 2. Click in graphics after each model is displayed. However. To understand how you determine whether a data set is isotropic or anisotropic. or an isotropic condition is said to exist when the rate of change of data values is the same in all directions. This is probably the most common case. an epithermal gold vein may have different rates of change in each of any three mutually perpendicular directions: along strike. anisotropy. The remainder of this chapter will deal with the use of isotropy and anisotropy in performing geostatistical estimations. you can see all values on the forms by ticking “Slow motion playback”. In geostatistical terms. you will need to study the chapters on variograms and variogram maps. isotropy. and how to calculate the direction and amount of anisotropy. or an anisotropic condition is said to exist when the rate of change of data values is different in different directions.tcl to see how these blocks are displayed in Surpac. the rate of change of alumina values in a large horizontal bauxite deposit beneath relatively flat topography may be isotropic in the XY plane. A true isotropic condition in three dimensions is rare for most types of data.Anisotropy Isotropy vs.

In the example below. In this string file. You will also assume that you are estimating a value at the block centroid (at coordinates 0N. the "weight" for each sample will be 0. one or more points.5 = 1. you will assume that all data is in the XY plane (i. the sample locations are represented by two points in a Surpac string file. Note that both samples are the same distance (3metres) from the block centroid. again assuming that all data is in the XY plane. in the image below. weight1 + weight2 = 0. For example. when the shape defining the line of equal weights is a circle. any sample whose location is on the circle shown below will be given the same weight as any other sample on that circle during the estimation of the value of the block centroid. where there are only two samples being used to estimate the value for the block. D1 contains the sample values (D1=10 for one point.Anisotropy Geostatistical Estimation Using Isotropy Geostatistical Estimation Using Isotropy In geostatistical estimation (inverse distance weighting. indicator kriging. are used to estimate a value at a location where there are no samples. Page 34 of 137 .5 + 0.5. etc.5 ) = 15 Throughout this tutorial. The location to be estimated is the centre position. and that only the two samples shown are going to be used for the estimation. you say that you have an "isotropic" condition.).0 When you assume that there is no directional continuity within the data. or "centroid" of a 1 x 1 x 1 block of material. the sample points and the block centroid all have the same Z value). In this example.5 ) + ( 20 * 0. ordinary kriging.e. you can assume that there is no "directional continuity" within the data. you are said to be performing an "isotropic" estimation.. Another way of stating this is that the "weight" applied to both samples will be equal. If you assume that the material surrounding the block and samples is homogenous (all the same). In this case. The calculation of the block value will be: ( sample value1 * weight1 ) + ( sample value2 * weight2 ) = block value ( 10 * 0. In two dimensions. you will assume that the sum of the weights must equal 1. and D1=20 for the other point). usually representing sample locations. In other words. and the two samples will contribute equally to the estimation. 0E).

25 + 0. and that only the distance from the sample to the block centroid is important.Anisotropy Geostatistical Estimation Using Isotropy This means that you are assuming that the direction from the point being estimated to the sample is not important. The calculation of the block value will be: ( 5 * 0.25 + 0. any samples falling on the surface of the same sphere will be given equal weight.25 ) + ( 35 * 0. all samples will be given equal weight.25 ) + ( 20 * 0. the sum of all the weights must be equal to 1.0 In three dimensions.5 As mentioned before. Page 35 of 137 . since all sample locations are the same distance from the block centroid. during isotropic estimation.25 ) + ( 10 * 0. In the example above.25 ) = 17.0: 0.25 + 0.25 = 1.

and are thus the same distance from the block centroid. you will be prompted to fill in values defining the orientation of the "major axis" and the "anisotropy ratios". For now. all samples will be given equal weight. when you are performing an estimation. use the following values: BEARING OF MAJOR AXIS: PLUNGE OF MAJOR AXIS: DIP OF SEMI-MAJOR AXIS: MAJOR/SEMI-MAJOR ANISOTROPY RATIO: MAJOR/MINOR ANISOTROPY RATIO: 0 (or any value from 0 to 360) 0 (or any value from -90 to 90) 0 (or any value from -90 to 90) 1 1 Page 36 of 137 .333 Again. all sample locations are on the surface of the same sphere.0 In Surpac.333 ) = 23. In this three-dimensional example of an isotropic condition.Anisotropy Geostatistical Estimation Using Isotropy In the example above.333 + 0.333 ) + ( 20 * 0.333 = 0.0 (assuming that 1/3 + 1/3 + 1/3 expressed as decimals equals 1): 0.333 ) + ( 40 * 0. if you wish to perform an estimation assuming that the data is isotropic. You will cover these topics later.999 = 1.333 + 0. The calculation of the block value will be: ( 10 * 0. the sum of all the weights is 1.

Anisotropy Geostatistical Estimation Using Isotropy To view an example of an isotropic sphere: 1. and then click Apply to display the D1 values for string 1 The concepts of “major axis”. 3. 2. Open isotropic_ellipsoid1. Page 37 of 137 . just understand that the lengths of all of these axes are the same for an isotropic ellipsoid. Choose Display > Point > Attribute. Enter the information as shown.str. “semi-major axis” and “minor axis” will be covered later. For now.

you will assume that the point being estimated is the centroid of the block. will have their anisotropic distances calculated as the actual distance (1. When a data set has anisotropy. Anisotropic conditions can result from geological conditions. and the sample whose value is 20 is 3metres from the block centroid. or similarity. an anisotropic condition is said to exist when the rate of change of data values is different in different directions. the two samples would be given the same weight in this case. and that only two samples. of samples taken along strike in a gold-bearing quartz vein may be better than the correlation of samples taken across strike. the correlation. in plan view. In this case. samples may be better correlated within the horizontal plane than vertically through the seam. such as fracturing. In a sedimentary deposit. This is commonly written as "a 3:1 anisotropy ratio". the anisotropic distance calculated for the sample whose value is 10 will be: Actual Distance x Anisotropy Ratio 1 x 3 = Anisotropic Distance = 3 Page 38 of 137 . are to be used to estimate a value for the block. dip: 90 degrees) a geologist may say that "there's about 3 times more continuity along strike (horizontally) than across strike (horizontally)". The direction of maximum continuity is referred to as the "major axis". Thus.Anisotropy Geostatistical Estimation Using Anisotropy Geostatistical Estimation Using Anisotropy As previously stated. many times this type of qualitative judgement is actually used in geostatistical estimation. as shown above. This is the case for nearly all data sets which represent samples taken from the earth. In this example. the direction from the point being estimated to a sample location is important. and not actual distances. How much anisotropy is present is also important. such as shown below: When you want to use anisotropy during an estimation. after becoming familiar with a silver deposit consisting of a vertical vein trending east to west (strike: 90 degrees. Even though the sample whose value is 10 is 1metre from the block centroid. deposition method. The determination of the magnitude of anisotropy for a data set may be done qualitatively or quantitatively (by intuition or by numerical calculation). the major axis could be defined as a bearing of either 90 or 270 degrees . As rough and unsubstantiated a statement as this may seem. Samples oriented due north or south of the block. For example. you can represent a 3:1 anisotropy ratio with a major axis bearing 90 degrees with an ellipse. In the silver vein example. such as the sample whose value is 10. Recall that you have indicated that there is a 3:1 anisotropy ratio and the bearing of the major axis is 90 degrees. such as a flat-lying coal seam. For example. This is because "anisotropic distances" are used in the calculation of the weights.they are both the same in geostatistical terms. you would say that there is a "3 to 1 anisotropy ratio" in the horizontal plane.. etc. in this case) multiplied by the anisotropy ratio (3. the direction from the location being estimated to the sample is important. in this case). In two dimensions.

75 Page 39 of 137 . as shown below. Sample Value Sample Bearing Actual Distance Anisotropy Factor Anisotropic Distance Weight 10 20 0 90 3 3 3 1 9 3 0.5 0. the weights assigned to both samples will change. The anisotropic distance of the sample whose value is 10 will now be 9: Actual Distance (3) X Anisotropy Ratio (3) = Anisotropic Distance (9).5 Since the anisotropic distances are the same.5 ) = 15 If the sample whose value is 10 is moved to a position at Y=3.5 ) + ( 20 * 0. X=0.Anisotropy Geostatistical Estimation Using Anisotropy This calculation is displayed in the following table for both samples: Sample Value Sample Bearing Actual Distance Anisotropy Factor Anisotropic Distance Weight 10 20 0 90 1 3 3 1 3 3 0. the weights for the points will be the same.25 0. and you again use a 3:1 anisotropy ratio with the bearing of the major axis at 90 degrees (or 270 degrees). The calculation of the block value will be: ( 10 * 0. This calculation is displayed in the following table for both samples.

In the example below. The calculation of the block value will now be: ( 10 * 0.5 Note that the calculation of the weights here is only approximate to demonstrate the effects of anisotropy. are all considered to be the same anisotropic distance from the block centroid: Page 40 of 137 . Assuming that our geologist has another opinion that "there is about 2 times more continuity horizontally along strike than vertically (up and down) within the plane of the vein". and the minor axis is the shortest. the semi-major axis is the second longest.75 ) = 17. and so. In two dimensions. So now you have a 3:1 anisotropy ratio in the horizontal XY plane. In actual practice. you would say that there is a "2:1 anisotropy ratio" in the vertical YZ plane.25 ) + ( 20 * 0. When you perform an estimation. In three dimensions. The ratio between the length of the major axis and the length of the semi-major axis is defined as the MAJOR/SEMI-MAJOR ANISOTROPY RATIO. Also. The ratio between the length of the major axis and the length of the minor axis is defined as the MAJOR/MINOR ANISOTROPY RATIO. the major axis is the longest.Anisotropy Geostatistical Estimation Using Anisotropy The weights of the samples will now be changed to reflect the new anisotropic distances. any samples falling on the surface of the same ellipsoid will be given equal weight. You distinguish between these ratios by defining three axes for the ellipsoid: Major axis Semi-major axis Minor axis By definition. and want to use three-dimensional anisotropy. this shape is called an "ellipsoid". an ellipse represents the line where weights are equal. the geostatistical method you decide to use will impact the values of the weights. all sample locations are on the surface of the same ellipsoid. all three axes are mutually perpendicular to one another. and a 2:1 anisotropy ratio in the vertical YZ plane.

333 ) + ( 25 * 0.5 2 3 0.333 ) = 13.333 + 0. the calculation of the weights for the data as shown will be: Axis Sample Value 5 Sample Bearing 90 Sample Dip 0 Actual Distance 3 Anisotropy Factor 1 Anisotropic Distance 3 Weight Major SemiMajor Minor 0.0 Page 41 of 137 .333 Since the anisotropic distances are the same. the sum of all the weights is 1.333 = 0. The calculation of the block value will be: ( 5 * 0.Anisotropy Geostatistical Estimation Using Anisotropy With the axes oriented as above.999 = 1.0 (assuming that 1/3 + 1/3 + 1/3 expressed as decimals equals 1): 0.333 25 0 90 1 3 3 0.3333 Again. the weights for the points will be the same.333 10 180 0 1.333 ) + ( 10 * 0. and a major/minor anisotropy ratio of 3. as well as a major/semi-major anisotropy ratio of 2.333 + 0.

Anisotropy

Geostatistical Estimation Using Anisotropy

If the distance from the block centroid to each sample is now the same, the weights will change. For example, in the view below, the distance from each sample to the block centroid is now 3, but you are still using the same anisotropy ellipsoid:

The calculation of the weights will be as follows:

Axis

Sample Value 5

Sample Bearing 90

Sample Dip 0

Actual Distance 3

Anisotropy Factor 1

Anisotropic Distance 3

Weight

Major SemiMajor Minor

0.5

10

180

0

3

2

6

0.333

25

0

90

3

3

9

0.1666

The calculation of the block value will be:

( 5 * 0.5 ) + ( 10 * 0.333 ) + ( 25 * 0.1666 ) = 7.75

Again, the sum of all the weights is 1.0 (assuming that 1/2 + 1/3 + 1/6 expressed as decimals equals 1):

0.5 + 0.333 + 0.1666 = 0.999 = 1.0

Page 42 of 137

Anisotropy

Ellipsoid Visualiser

Ellipsoid Visualiser

Using our previous example where you have a major/minor anisotropy ratio of 3, and a major/semi-major anisotropy ratio of 2, you would get an ellipsoid, but you need to establish the orientation of the ellipsoid. In Surpac, this can be accomplished in several different ways, including the “Surpac” method. The examples which follow use the “Surpac” method, which encompasses the following three terms:

Term Bearing of major axis Plunge of major axis

Min 0 -90

Max

Description

360 azimuth of major axis in XY plane 90 dip above or below horizontal plane rotation of semi-major axis around major axis

Dip of semi-major axis

-90

90

The ellipsoid visualiser is a tool which can assist you to understand the orientation of the anisotropy ellipsoid. You will now use it to create several anisotropy ellipsoids, and save them as Surpac string files.

1.

Choose Geostatistics > Ellipsoid visualiser.

The form below is displayed:

Page 43 of 137

Anisotropy

Ellipsoid Visualiser

You can use the values of bearing, plunge, and dip in the following examples to create the ellipsoids in each example.

Example #1:

This ellipsoid could be used to estimate gold values within a vertical vein that has strike: 90 degrees and dip: 90 degrees.

Bearing of major axis Plunge of major axis Dip of semi-major axis Major/semi-major anisotropy ratio Major/minor anisotropy ratio

90 0 -90 2 3

Looking down on XY plane 3D View major/minor anisotropy ratio: 3

Looking north at XZ plane major/semi-major anisotropy ratio: 2

Looking west at YZ plane dip of semi-major axis: 90

Page 44 of 137

2. Choose Display > 3D Grid. and then click Apply.Anisotropy Ellipsoid Visualiser To view an example of this anisotropic ellipsoid: 1. and then click Apply to display D1 values for string 1 Page 45 of 137 . Choose Display > Point > Attributes. Enter the information as shown. 5. 3.str. Enter the information as shown. 4. Open anisotropic_ellipsoid1.

but the continuity is significantly less in the vertical direction. Bearing of major axis Plunge of major axis Dip of semi-major axis Major/semi-major anisotropy ratio Major/minor anisotropy ratio 0 0 0 1 5 Looking down on XY plane 3D View major/semi-major anisotropy ratio: 1 Looking north at XZ plane dip of semi-major axis: 0 Looking west at YZ plane major/minor anisotropy ratio: 5 Page 46 of 137 . where continuity within the seam is the same in the XY plane (major/semi-major anisotropy ratio: 1).Anisotropy Ellipsoid Visualiser Example #2: This ellipsoid could be used to estimate values within a horizontal coal seam or other data from flat-lying sedimentary rocks.

Enter the information as shown. Open anisotropic_ellipsoid2. and then click Apply to display D1 values for string 1 Page 47 of 137 .Anisotropy Ellipsoid Visualiser To view an example of this anisotropic ellipsoid: 1. 3. Enter the information as shown. Choose Display > Point > Attributes. 5. 2. and then click Apply.str in graphics Choose Display > 3D Grid. 4.

Bearing of major axis Plunge of major axis Dip of semi-major axis Major/semi-major anisotropy ratio Major/minor anisotropy ratio 180 -60 0 3 3 Looking down on XY plane 3D View major/semi-major anisotropy ratio: 3 Looking north at XZ plane dip of semi-major axis: 0 Looking west at YZ plane major/minor anisotropy ratio: 3 Page 48 of 137 . which plunges to the south at a dip of 60 degrees below the horizontal.Anisotropy Ellipsoid Visualiser Example #3: This ellipsoid could be used to estimate values from a kimberlitic diatreme. or diamond-bearing "pipe" type ore body.

Page 49 of 137 .Anisotropy Ellipsoid Visualiser To view an example of this anisotropic ellipsoid: 1.str in graphics. and then click Apply. Choose Display > 3D Grid. Enter the information as shown. 4. Open anisotropic_ellipsoid3. 3. 2. Enter the information as shown. 5. and then click Apply to display D1 values for string 1. Choose Display > Point > Attributes.

Bearing of major axis Plunge of major axis Dip of semi-major axis Major/semi-major anisotropy ratio Major/minor anisotropy ratio 50 0 -60 1 3 3D View Looking down on XY plane Looking north at XZ plane Looking west at YZ plane Page 50 of 137 . with strike of 50 degrees and dip to the southeast of 60 degrees below the horizontal. where continuity within the vein is the same in all directions (major/semi-major anisotropy ratio: 1).Anisotropy Ellipsoid Visualiser Example #4: This ellipsoid could be used to estimate values from an epithermal vein.

Anisotropy Ellipsoid Visualiser Example #4 (continued): Looking horizontally along strike: az 50 degrees. Enter the information as shown. Choose Display > 3D Grid. and then click Apply. Open anisotropic_ellipsoid4. dip 0 Note dip of semi-major axis is -60 degrees major/semi-major anisotropy ratio: 1 Looking downdip: azimuth 140 degrees. dip -60 note major axis is along strike. semi-major is downdip major/minor anisotropy ratio: 3 To view an example of this anisotropic ellipsoid: 1. 4. Choose Display > Point > Attributes. 3. 2. Page 51 of 137 .str.

Anisotropy Ellipsoid Visualiser 5. and/or experience with several data sets to grasp the concepts presented here. It may take some time. and then click Apply to display D1 values for string 1. you should understand how anisotropy ratios and orientation of the anisotropy ellipsoid impacts the calculation of anisotropic distances. Page 52 of 137 . Summary You should now understand the following terms: Isotropy Anisotropy Anisotropic ellipsoid Major axis Semi-major axis Minor axis Major/Semi-major anisotropy ratio Major/Minor anisotropy ratio Anisotropic distance Sample weight Also. and therefore the weight used for samples in estimating a value at a block centroid. more research. Enter the information as shown. Understanding and visualising an anisotropy ellipsoid and how it impacts upon an estimation is no simple task.

Omnidirectional variograms. 6. 3. you should: • • • be familiar with Surpac string files. 7. Directional variograms. 5. Modifying the lag distance. Calculating an omnidirectional variogram in Surpac. Requirements Prior to proceeding with this chapter. Calculating a variogram. Introduction to the variogram. 4. Modelling variograms in Surpac. know how to run a Surpac macro. 2. understand basic statistical concepts such as mean and variance. A variogram is a graphical tool which can be used to describe these concepts. The variogram will be described through the following: 1.Variograms Ellipsoid Visualiser Variograms Overview An important aspect of performing any geostatistical evaluation is to understand how data values change over distance and direction. Page 53 of 137 .

while the other half does not contain any gold. Although these areas are beyond the scope of this tutorial. and during data entry. you should be aware of them. there is some difference. and send it to two different labs. The nugget value is noted as a difference at a sample separation distance of zero: The Nugget The term “nugget” comes from a situation that often occurs in coarse gold deposits where a sample is split. This difference is called the “nugget”. in the lab.Variograms Introduction to the Variogram Introduction to the Variogram A variogram is a graph that compares differences between samples against distance: The Variogram Nugget If you split a single sample. and one half contains a gold nugget. very often you will get two different values. and their impact on the nugget and subsequent geostatistical evaluations. Page 54 of 137 . at a sample separation distance of zero. human error can also be a factor. Any or all of these can contribute to the nugget. Thus. Although differences between sample “splits” is often responsible for this. Errors occur in sampling. also abbreviated as “c(0)”.

Beyond the range. For example. and is calculated as: nugget to sill ratio = nugget / (nugget + sill) Range The distance at which the sill is attained is referred to as the range: The Range The range (sometimes abbreviated as the letter “A”) represents the maximum distance which sample pairs can be said to have some relationship to their separation distance. you would expect the difference to be greater than samples which are closer together.Variograms Introduction to the Variogram Sill If you compare two samples some distance apart. Page 55 of 137 . this maximum difference is displayed as the flat portion of the graph. On the variogram. as shown above. the difference between the samples cannot get any greater. At some point. is the difference between the maximum difference and the nugget. The portion of the graph of the variogram which rises up and to the right of the nugget point represents this situation. there is no relationship. The term “nugget to sill ratio” is used to describe what percentage of the “total sill” the nugget comprises. Two values describe the point at which the variogram reaches its maximum value – the sill and the range. the maximum sample value minus the minimum sample value gives us the greatest difference between samples. The Sill The sill (sometimes abbreviated as the letter “C”).

a data set is grouped into “pairs”. 3-4. you get the data pairs of 3-3. Assume that the values represent samples taken at 1 metre intervals along a north – south line: 3 3 4 6 7 5 5 3 To create the variogram graph of “Distance vs. you will use the data below. For the first lag interval of 1. or “lag”. which are separated by a given distance. or “lag interval”.0 Page 56 of 137 . you first decide upon a lag distance. then the following calculation is performed on all samples in each bin: sum of (difference between sample values)2 gamma(h) = 2 x number of pairs To demonstrate this. Difference”. You then group the data into sample pairs which fall into each lag interval. etc… The difference between the two values is squared.Variograms Calculating a Variogram Calculating a Variogram To calculate a variogram. 4-6. Then. and the sum of all squared distances is calculated: Lag = 1 Pair 1 2 3 4 5 6 7 Pair Values 3–3 3-4 4-6 6-7 7-5 5-5 5–3 Difference 0 -1 -2 -1 2 0 2 Squared difference 0 1 4 1 4 0 4 Sum of squared differences: 14 sum of squared differences gamma(h) = 2 x number of pairs = 14 2x7 = 1.

all samples separated by lag distances of 2 are paired off.Variograms Calculating a Variogram Next.3 The results of lag distances of 3. and 5 are below: Lag = 3 Pair 1 2 3 4 5 Pair Values 3–6 3-7 4-5 6-5 7–3 Difference -3 -4 -1 1 4 Squared difference 9 16 1 1 16 Sum of squared differences: 43 sum of squared differences gamma(h) = 2 x number of pairs = 43 2x5 = 4. 4. and the calculation is performed again: Lag = 2 Pair 1 2 3 4 5 6 Pair Values 3–4 3-6 4-7 6-5 7-5 5–3 Difference -1 -3 -3 1 2 2 Squared difference 1 9 9 1 4 4 Sum of squared differences: 28 sum of squared differences gamma(h) = 2 x number of pairs = 28 2x6 = 2.3 Page 57 of 137 .

8 Lag = 5 Pair 1 2 3 Pair Values 3–5 3-5 4–3 Difference -2 -2 1 Squared difference 4 4 1 Sum of squared differences: 9 sum of squared differences gamma(h) = 2 x number of pairs = 9 2x3 = 1.3 4.5 Page 58 of 137 .8 1.3 3.Variograms Calculating a Variogram Lag = 4 Pair 1 2 3 4 Pair Values 3–7 3-5 4-5 6–3 Difference -4 -2 -1 3 Squared difference 16 4 1 9 Sum of squared differences: 30 sum of squared differences gamma(h) = 2 x number of pairs = 30 2x4 = 3.5 All of the results and lag distances are then compiled: Lag (distance) 1 2 3 4 5 gamma(h) (difference) 1 2.

2 Sill: 4.2+4. increases by the sill at a point defined by the range.05 Page 59 of 137 . The variogram model can be described by a mathematical equation. all relevant parameters of the model would be recorded: Nugget: 0.Variograms Calculating a Variogram A graph of the results looks like this: Experimental Variogram This graph of calculated gamma(h) values versus lag distance is referred to as an “experimental variogram”.0) = 0. A variogram model. as shown in the previous section.2 / (0. starts at a nugget. A variogram model has been fit to the experimental variogram below: Variogram Model fitted to Experimental Variogram In this example. This is used to calculate the variogram displayed in the previous version . then continues infinitely to the right at the total sill value.0 Nugget/Sill ratio = 0. and is subject to the interpretation of the person who is analysing the data.a “variogram model”.0 Range: 3.

To demonstrate the impact of the value chosen for the lag.001 and 4. representing samples taken at 1metre intervals along a north – south line: 3 3 4 6 7 5 5 3 For the 0-2 lag bin. you now get the data pairs below: Lag = 2 Pair 1 2 3 4 5 6 7 8 9 10 11 12 13 Pair Values 3-3 3-4 3-4 3-6 4-6 4-7 6-7 6-5 7-5 7-5 5-5 5-3 5-3 Difference 0 -1 -1 -3 -2 -3 -1 1 2 2 0 2 2 Squared difference 0 1 1 9 4 9 1 1 4 4 0 4 4 Page 60 of 137 . Here again is the data. but using a lag interval of 2.001 and 3. “Lag=2” meant “all sample pairs whose separation distance is between 1. Lag=6 sample pairs whose separation distance is between 4. you will recalculate the variogram in the previous example. In the previous example. a lag interval of 1 was used. Lag=3” meant “all sample pairs whose separation distance is between 2.001 and 1.Variograms Modifying the Lag Distance Modifying the Lag Distance Although the previous example generated a well-formed experimental variogram. The term “Lag=1” actually meant “all sample pairs whose separation distance is between 0. You will calculate three “lag bins”: Lag=2 sample pairs whose separation distance is between 0 and 2 Lag=4 sample pairs whose separation distance is between 2.001 and 6.001 and 2“. often it is necessary to modify the lag distance to obtain such a good-looking variogram.

3 . you now get the data pairs below: Lag = 4 Pair 1 2 3 4 5 6 7 8 9 Pair Values 3-6 3-7 3-7 3-5 4-5 4-5 6-5 6-3 7-3 Difference -3 -4 -4 -2 -1 -1 1 3 4 Squared difference 9 16 16 4 1 1 1 9 16 Sum of squared differences: 73 sum of squared differences gamma(h) = 2 x number of pairs = 73 2x9 = 4.6 For the 2-4 lag bin.1 Lag = 6 Pair 1 2 3 4 5 Pair Values 3-5 3-5 3-5 4-3 3-3 Difference -2 -2 -2 1 0 Squared difference 4 4 4 1 0 Sum of squared differences: 13 sum of squared differences gamma(h) = 2 x number of pairs = Page 61 of 137 13 2x5 = 1.Variograms Modifying the Lag Distance Sum of squared differences: 42 sum of squared differences gamma(h) = 2 x number of pairs = 42 2x13 = 1.

Variograms Modifying the Lag Distance All of the results and lag distances are then compiled and graphed: Lag (distance) 2 4 6 gamma(h) (difference) 1.6 4.1 1.3 Experimental Variograms with lags of 1(solid) and 2(dashed) Page 62 of 137 .

In three dimensions. Move to the first point. Determine which points not selected are within the second lag tolerance distance from this point. Remove the previous point from consideration. The example below demonstrates how sample pairs would be selected for a data set. Page 63 of 137 . and 3 are used. In an omnidirectional variogram. sample pair 1-4 is oriented north-south. Determine which other points in the data set are within the first lag tolerance distance from this point. the orientation of the sample pairs is irrelevant. All samples used were aligned north-south. Repeat steps B to F until all points have been considered. the pairs are selected based only on their separation distance. e. b. In this type. d. Repeat until all points have been put into a lag bin. and not on the orientation of the pairs. and add these pairs to the first “lag bin” (Lag=1). and lag values of 1. sample pair 1-2 is oriented east-west. the search from each point takes the shape of a sphere. Another type of variogram is known as an “omnidirectional variogram”. 2. The manner which the software determines pairs is this: a. Move to the next point. g. and yet both pairs are used for the “Lag=1” bin.Variograms Omnidirectional Variograms Omnidirectional Variograms The variogram in the previous exercise was an example of a “directional” variogram. c. All samples are on a 1x1 grid. Lag selection circles Sample pairs selected for each lag in an omnidirectional variogram Lag=1 1-2 1-4 2-3 2-5 3-6 4-5 5-6 Lag=2 1-3 1-5 2-4 2-6 3-5 4-6 Lag=3 1-6 3-4 Note: The example here is two-dimensional. f. For example. and add them to the second “lag bin” (Lag=2).

using a northeast – southwest orientation of 45 degrees. Repeat steps B to F until all points have been considered. and add them to the second “lag bin” (Lag=2). 3-4. and add these pairs to the first “lag bin” (Lag=1). all samples were aligned north-south. e. Remove the previous point from consideration. and 3 are used. so the only variogram possible was a directional variogram. Lag selection circles with directional tolerance search Page 64 of 137 . 2. b. The example below demonstrates how sample pairs would be selected for a data set. Determine which points not yet selected are within the second lag tolerance distance from this point AND within the angular tolerances. In the first example. plus or minus some angular tolerance. In Surpac.5 degrees either side of that direction. and 3-5. These are the data pairs: 2-4. The manner in which the software determines pairs is shown below: a. this angular tolerance is known as the spread. However.5 degrees. but neither are within the angular tolerance of 45 degrees plus or minus 22. Thus. sample pairs 1-2 and 1-4 are both within the first lag tolerance. Move to the first point. In a directional variogram. In a directional variogram. plus or minus a spread tolerance of 22. in most data sets. it will be included in the calculation.5 and 67. In fact. the software selects only those data pairs which are oriented in a particular manner. there are a multitude of data pair orientations. c. For example. the orientation of the sample pairs is important. Move to the next point. f.5 degrees). There was no other possible orientation for the sample pairs to take. d. All samples are on a 1x1 grid. there are only three data pairs in the entire data set which have an orientation that is within the defined limits.5 degrees (or 202.Variograms Directional Variograms Directional Variograms A directional variogram is one in which all sample pairs are oriented in a particular direction. Determine which other points in the data set are within the first lag tolerance distance from this point AND within the angular tolerances. if a sample pair is oriented between 22. and lag values of 1.5 and 247.

the concept of a “directional” variogram could be questioned. If a tolerance angle is too small. the quality of the experimental variogram may be reduced to the point that a model cannot be fitted with any confidence. Surpac has the option to restrict the radius of the cone to a maximum via the use of a “spread limit”. Page 65 of 137 . the search from each point takes the shape of a cone.Variograms Directional Variograms Sample pairs selected for each lag for a directional variogram (orientation 45 +/. In three dimensions. so does the number of pairs. This has the effect of turning the search cone into a cylinder with the radius of the spread limit. If the tolerance angle is too large. As the tolerance angle decreases. using directional variograms reduces the number of sample pairs.22. Note that the example here is two-dimensional.5) Lag=1 Lag=2 2-4 3-5 Lag=3 3-4 As you can see. Additionally.

Since it will usually be the 'best' variogram. An example of a directional variogram is one in which only those sample pairs which are due north or south of one another are used. and all sample pairs are used. An omnidirectional variogram is defined by setting the spread angle to 90. an omnidirectional variogram is a good place to start. Calculating an omnidirectional variogram Page 66 of 137 . the omnidirectional variogram will be the 'best' or 'least scattered' variogram.tcl. Directional variograms only consider sample pairs which are aligned within a certain orientation. Variograms can be directional or omnidirectional. Next. The macro will open the VARIOGRAM MODELLING window by choosing Geostatistics >Variogram modelling. as it will contain more data points than a directional variogram. The azimuth and plunge values are irrelevant. After reading the text below on the first form. Run the macro 04_omnidirectional_variogram.Variograms Calculating an Omnidirectional Variogram in Surpac Calculating an Omnidirectional Variogram in Surpac 1. An omnidirectional variogram is one in which sample orientation is not considered. 2. the macro chooses Variogram > New string file variogram and the form below is displayed. click Apply. Generally.

The Spread is the angular tolerance in decimal degrees either side of the search axis.not will be created. In practice. This will ensure that you do not miss any long-range effects. When you are creating your first omnidirectional variogram. as a spread angle of 90 degrees includes all data. The D1 field contains the gold values. A maximum distance of 100metres indicates that sample pairs further than this separation distance are not to be considered. An initial lag distance of 2metres is chosen to illustrate a point later in this exercise. Page 67 of 137 . the “spread cone” becomes a cylinder. X. or Z coordinate values on the Advanced tab.Variograms Calculating an Omnidirectional Variogram in Surpac You have nominated to use all strings (by leaving the string range blank) from the file gold_cut17. The Plunge is the dip above (positive) or below (negative) the horizontal. The Azimuth is the orientation of the search axis in the XY plane. You can zoom in on a portion of the variogram afterward.str to calculate the omnidirectional variogram. After it reaches the Spread limit. The values of Azimuth and dip are set to zero in this case. you may want to select an initial lag value for your first variogram to somewhere near the sample spacing. a Spread Limit may be specified. or. you may want to set this to a fairly large value – even near the maximum extent of your data. All variograms in Surpac are defined by the three fields of Azimuth. You can exclude data which is greater or less than any Y. but could be anything. The output report file omnidirectional. cut to a maximum value of 17 grams per tonne. if you recalculate it. Optionally. The Search Cone The way in which you specify an omnidirectional variogram in Surpac is to use a spread angle of 90 degrees. and Spread. The Spread limit is an optional maximum distance from the search axis which the “spread cone” is allowed to be expanded to. The fields “Minimum value” and “Maximum value” allow you to exclude data which is below a given minimum value or above a given maximum value. regardless of the azimuth or dip. Plunge. you can use a smaller maximum distance to decrease processing time.

The next form will set the lag to 7. After reading the text below on the form. Note that these buttons will modify the lag by the increment specified on the Advanced tab of the first variogram modelling form.Experimental variogram lag.1 Page 68 of 137 . which will result in a 'better' variogram. Note that the initial lag value of 2 has created a poor quality or 'scattered' variogram.6. click Apply. you can set the minimum. Drag the lag slider at the top right corner of the window. and incremental values of the lag slider. 4. Use the < and > buttons on either side of the lag slider. After this macro is completed. Select Variogram . The Lag Slider 4.Variograms Calculating an Omnidirectional Variogram in Surpac 3. you can adjust the lag by a value of 0. Select the Advanced tab. Click Apply on the form. 2. enter a value and Apply the form. Additionally.1 using the right and left arrow keys on your keyboard. You can see the lag slider to the right of the flashing macro playback button on the variogram modelling window. On the Advanced tab. This tool allows you to interactively modify the lag and see the effects instantly on the variogram. you can modify the lag value by one of the following methods to see what effect the lag value has on the quality of the variogram: 1. 3. Use the right and left arrow keys on the keyboard to modify the lag by 0. 5. The lag distance can be modified using a tool called the “lag slider”. maximum.

A form will also display. The experimental variogram will be modified to reflect the new lag distance. This was displayed by selecting Display > Display/Hide variance. Next. Click Apply to accept the lag value of 7.6. The form to specify a value for the variogram lag is presented: Setting the lag to 7. Many geostatisticians model the variogram so that the “total sill” is equal to the variance.Variograms Calculating an Omnidirectional Variogram in Surpac You will see that the 2 metre lag results in a scattered variogram: Experimental variogram with lag=2 The value of the variance of the data set is represented by a horizontal green line. Page 69 of 137 . You do not have to use the variance if you do not want to.6 6. the macro will choose Variogram > Experimental variogram lag.

Surpac attempts to “auto-fit” the variogram to the experimental variogram. or may be defined exactly. 9. 8. and one or more sill-range structures. You may click and drag the cyan (light blue) dots to modify the shape of the variogram. click Apply. The range is the distance along the X axis value for a given sill value. The sill is the Y axis distance above the nugget for a structure.Variograms Calculating an Omnidirectional Variogram in Surpac 7. When Variogram > Model is invoked for the first time.6 Additionally. Click and drag the form away so you can see the new experimental variogram. The nugget and sill-range structures may be modified graphically. Page 70 of 137 . After reading the text below on the form. shown as the thin red line. the macro selected Variogram > Model to display the variogram model. Up to five sill-range structures can be used. but often one will be enough. Experimental variogram with lag set to 7. The nugget is the value of the model on the Y axis. as on the following form. A variogram model is made up of a nugget.

String file output of variogram data This file can be used in the Plotting module to generate hardcopy output. Leave the form as is. It is a good idea to save the omnidirectional variogram model parameters for future reference. The output file omnidirectional_variogram.Variograms Calculating an Omnidirectional Variogram in Surpac Next.str containing the experimental variogram and the variogram model by choosing File > Save as > String file.not will be created. and will contain all of these values. and click Apply. the macro chooses Variogram > Report to display the variogram model parameters: Variogram model parameters You may also explicitly type in values on this form when modelling your own variograms. 10. This file is displayed in graphics. The macro saves the string file omnidirectional1. Page 71 of 137 .

3. The Y axis has been set to 5x normal exaggeration using View > Data view options > View scale factors The experimental variogram has been renumbered from string 1 to string 30003 using Edit > String > Renumber range. 2. and closed the variogram modelling window. String 30003 has a default presentation style of a small round marker The real world X value has been displayed on string 2 to represent the values you see from 0 to 100 on the X axis via the use of Display > Point > X values The real world Y value has been displayed on string 3 to represent the values you see from 0 to 12 on the Y axis via the use of Display > Point > Y values The description field on string 5 has been modified to include the words in brackets using Edit > Point > Properties The macro has now completed.Variograms Calculating an Omnidirectional Variogram in Surpac Note that a few modifications have been performed on the string file to present the view you see: 1. 5. 4. Page 72 of 137 .

variogram model . After that. the following files have been saved: omnidirectional. Surpac allows you to create a file containing the experimental variogram data as well as a file containing a variogram model. Experimental variogram opened from *. Choose Display > Display/Hide variance. then open and modify the previous example’s variogram model.EVG and . From running the previous example. The variance can be used as the “total sill” of the variogram.evg. you will open the experimental variogram.VGM files in navigator Next.EVG file 2. Page 73 of 137 . 1. you will create a new model. You may or may not choose to use the variance for modelling.Variograms Modelling Variograms in Surpac Modelling Variograms in Surpac In order to make it possible to review or re-model a previously calculated experimental variogram. Open omnidirectional.experimental variogram omnidirectional. The variance of all data points used to calculate the variogram will be displayed as a horizontal green line.vgm .evg .

Variograms Modelling Variograms in Surpac 3. and variance There are several features to assist you with modelling a variogram: 4. Practice using the lag slider at the upper right portion of the variogram window. drag and drop the file omnidirectional.evg file is opened first. Recall that the maximum. To review a previously created variogram model.1 The Lag Slider Page 74 of 137 . Experimental variogram. minimum.vgm into the variogram window. and that the right and left arrow keys on your keyboard will increment and decrement the lag by 0. Note: You cannot open a *. and incremental values were set on the Advanced tab. variogram model.vgm file unless the *.

When you move this marker. You will now modify the existing variogram model. the shape of the variogram will change. Number of pairs The larger the number of pairs. You can modify the nugget with the leftmost marker. Choose Display > Set axis limits. and the entire variogram will move up and down. you will see that it will remain on the Y axis. As you move this marker.Variograms Modelling Variograms in Surpac 5. the more confidence you can have in a point on the experimental variogram. Set the parameters as shown. but the nugget will remain the same. You can modify the sill and range with the rightmost marker. and click Apply: Set variogram axis limits This feature can be useful to “zoom in” on a portion of the variogram. Choose Variogram > Model. Page 75 of 137 . 6. Some geostatisticians tend to ignore certain experimental data points if they have a substantially lower number of pairs than the surrounding data. 9. Choose Display > Display/Hide Number of pairs. As you can see in this example. Click and drag either of the cyan (light blue) coloured markers. 7. 8. the first experimental variogram point has been ignored.

Automatic fit variogram The software uses an algorithm which attempts to fit a variogram model to the data presented. As you click and drag the markers to modify the variogram. 13. 10. you will delete the model and create a new one.evg Choose Display > Display/Hide variance.Variograms Modelling Variograms in Surpac Next. 11. Choose Variogram > Delete model. Open omnidirectional. 14. notice that the current variogram parameters are displayed in the upper right part of the window: Page 76 of 137 . Choose Variogram > Model. This may or may not be close to your final result. You will be prompted to “select and drag a marker”. Move the cyan markers until you get a variogram similar to the previous example. 12.

Choose Variogram > Report. Enter the values below and click Apply. If you open the file in a text editor.not will be created in the current working directory. you will see the values stored there: Page 77 of 137 . 16.Variograms Modelling Variograms in Surpac Variogram parameters in upper right 15. Variogram report The file omni_report.

You should now have three markers – one nugget and two sill/range markers. Choose Variogram > Add structure again. 22. Enter the values shown on the form below (left) and then click Apply. Choose Variogram > Add structure.0000. Multiple structures Multiple structures As you can see. Click and drag each of the markers and note the effect. 21. Notice that one of the cyan markers will be removed. Click and drag each of the markers and note the effect. A third cyan marker will be displayed. Surpac allows up to five structures to be added.00048. The sum of the nugget and all sill values equals the “total sill”. 20. Page 78 of 137 . If multiple slopes are required to adequately model the data. but it also provides a way to set the variogram parameters to a specific value (a Sill of 4. additional “structures” can be added. 23. 19. and the model is modified. 18.NOT file which contains the variogram model parameters. One or two structures will generally be sufficient for modelling most experimental variograms. for example). 17.Variograms Modelling Variograms in Surpac Not only does the Variogram – Report function allow you to create a *. rather than 4. Choose Variogram > Report. Choose Variogram > Delete structure.

Variograms

Modelling Variograms in Surpac

Modelling a variogram is not an exact science. However, here are a few tips:

a.

b.

c. d.

Try to model the “trend”. One way to see the “trend” of the data is to create a variogram model (the red line), then drag the lag slider back and forth. The model should fit the data reasonably well for a few different lags. Consider the geology of the domain. Does the model you come up with look reasonable for your data? Is the nugget effect high or low? Is the range about what you expected? If not, perhaps you should reconsider the data domains. Search the Internet. There are many online resources with explanations of variograms, variogram calculations, and results of particular data sets. Ask a consultant. Although they are generally not free, geostatistical consultants are perhaps the best source of geostatistical advice.

Assuming that you are reasonably satisfied with your multiple-structure variogram, the next step is to save it.

24.

Choose File > Save > Experimental variogram and model.

Save Experimental variogram and model - menu

You may save either the experimental variogram or the variogram model alone, or you can, as you are doing in this case, save both of them at the same time. 25. Enter the information as shown, and then click Apply:

Saving experimental variogram and model

The following files will be created: multi_structs.evg - experimental variogram multi_structs.vgm - variogram model Next, you will demonstrate how these files can be used to re-display previous models.

Page 79 of 137

Variograms

Modelling Variograms in Surpac

26.

Choose Variogram > Delete model. This will remove the current model.

27.

Click and drag multi_structs.vgm into the variogram viewport. This will redisplay the multiple-structured model.

28.

Click and drag omnidirectional.evg to redisplay the original omnidirectional experimental variogram. Note that omnidirectional.evg and multi_structs.evg are identical, so no change in the experimental variogram will be apparent. However, the variogram model will be removed.

Now you will briefly look at the different “types” of variograms.

29. 30.

With the omnidirectional experimental variogram displayed, Choose Variogram > Type. Select Logarithmic, and then click Apply. Note that not only does the shape of the variogram change, but the gamma(h) values are also modified. If the variance of the data was displayed, it is now removed.

31. 32.

Once again, Choose Variogram > Type. Select General Relative and then click Apply. Note that the shape in this case is identical to the normal variogram, but the gamma (h) values are different. They have been divided by the value of the variance of the data set, which in this case was very close to the value of 10.

33. 34.

Once again, choose Variogram > Type. Select Pairwise Relative and then click Apply. Note that the shape in this case is different to the normal and general relative, and the gamma(h) values are different again.

Normal

Logarithmic

Page 80 of 137

Variograms

Modelling Variograms in Surpac

General Relative

Pairwise Relative

Logarithmic variogram In this type of a variogram, logarithms of the raw data are used to calculate the experimental variogram. Any negative or zero raw data values are first set to small positive values before a logarithmic transformation is applied. If the distribution of your data is logarithmic, or near logarithmic, and a reasonable fit cannot be obtained from a normal variogram, a logarithmic variogram can potentially give a usable value which can be used for the range of a normal variogram. If local values within the population are affected by the local variance, a condition called a “proportional effect” is said to exist. “Relative” variograms can be more useful than “normal” variograms if a proportional effect exists.

General Relative variogram In this type of a variogram, each gamma(h) is divided by the squared mean of all samples used to estimate that gamma(h).

Pairwise Relative variogram In this type of variogram, the squared difference of each sample pair is divided by the square of the mean of each sample pair. As for the logarithmic variogram, units along the Y (gamma(h)) axis have no meaning but the relative variograms may serve to identify ranges and structures

Note: The application of these types of variograms is an advanced topic. If you wish to obtain more information on these, you should obtain instruction from a qualified geostatistical consultant.

The last topic of discussion in this tutorial is that of spherical versus exponential models.

35. 36. 37.

Redisplay the normal variogram by dragging and dropping omnidirectional.evg into the variogram viewport. Now display the variogram model by dragging and dropping omnidirectional.vgm into the variogram viewport. Choose Display > Set axis limits.

Page 81 of 137

Variograms

Modelling Variograms in Surpac

38.

Set the parameters as shown, and then click Apply:

Set variogram axis limits 39. Choose Variogram > Spherical / Exponential model. Note the difference in the two models, as displayed below.

Spherical Model

Exponential Model

A spherical model has a much straighter increasing section, then at the range, the sill of the model becomes a constant. In an exponential model, the entire model is curved, and the sill is a value which the model approaches, but never actually attains. If your experimental variogram continually increases, then an exponential model may be a more accurate means of representing the data than a spherical model. If, however, as in the case above, the experimental variogram first increases and then decreases, a spherical model is probably a better representation of the data. The last menu item you will cover here is the experimental variogram report. 40. 41. Choose Variogram > Experimental variogram report. Enter the information as shown below, and then click Apply:

Experimental variogram report

Page 82 of 137

Variograms

Modelling Variograms in Surpac

**The following data will be present in the file exp_report.not:
**

Jun 26, 2007 Surpac Minex Group VARIOGRAM CALCULATION Data Source: gold_cut Id : 17 Strings : all Output File: exp_report.not

D Field : 1 Valid Data Range : All values Lag : 7.6 Max Distance : 100 VARIOGRAM DIRECTION Azimuth : 0.000 Plunge : 0.000 Spread angle : 90 Spread limit : None STATISTICS Number of samples : 335 Mean : 3.181744 Variance : 9.842981 Standard Deviation : 3.137353

Lag Pairs Drift Gamma(h) Wtd Gamma(h) Log Gamma(h) Gen Rel Gamma(h) P/w Rel Gamma(h) Avg Distance) ----------------------------------------------------------------------------------------------------------------0.00 7.60 15.20 22.80 30.40 271 -0.383 567 -0.600 176 -1.622 174 -0.369 228 -0.498 7.941 7.424 9.219 9.805 9.362 8.824 7.962 8.943 7.036 9.210 7.802 12.435 7.764 7.178 7.941 7.220 9.342 9.652 9.439 8.861 7.980 8.957 7.080 9.239 7.809 12.539 7.730 7.170 0.311 0.314 0.410 0.458 0.393 0.392 0.353 0.388 0.335 0.380 0.329 0.447 0.358 0.353 0.784 0.733 0.911 0.969 0.925 0.872 0.787 0.883 0.695 0.910 0.771 1.228 0.767 0.709 0.226 0.226 0.265 0.309 0.275 0.276 0.248 0.274 0.243 0.267 0.236 0.296 0.254 0.254 2.000 6.250 14.565 23.057 30.248 39.307 45.648 53.366 60.671 67.946 76.427 83.910 90.897 97.618

38.00 1155 0.554 45.60 1556 0.271 53.20 1626 0.176 60.80 1886 0.391 68.40 1474 0.167 76.00 1106 0.351 83.60 2041 0.442 91.20 2159 0.318 98.80 1415 0.700

Page 83 of 137

and save experimental variograms and variogram models in Surpac.Variograms Modelling Variograms in Surpac Summary You should now understand how a variogram is calculated. Page 84 of 137 . and how to create. view.

variogram maps can be used to define the anisotropy ellipsoid.Variogram Maps Modelling Variograms in Surpac Variogram Maps Overview An important aspect of performing any geostatistical evaluation is to understand the anisotropy of the data. or which direction has the longest continuity. know how to calculate and model a variogram in Surpac. 2. understand the parameters which define an anisotropy ellipsoid. understand the concept of an anisotropy ellipsoid. as well as the two mutually perpendicular directions. Secondary variogram map. 3. you should: • • • • be familiar with Surpac string files. These concepts are explained through the following: 1. Primary variogram map. Page 85 of 137 . Requirements Prior to proceeding with this chapter. and how data values change with regard to that direction. Calculation of anisotropy ellipsoid parameters. Additionally. A variogram map is a tool within Surpac which allows you to visualise anisotropy in a plane.

it is important to determine if there is an orientation of the data where there is more continuity than any other orientation. The following numeric values are required to define a plane for a variogram map: Dip (-90 to +90) and Dip direction (0 to 359) For example. click Apply.tcl. there is an orientation that has greater continuity than all others. an isotropic condition is said to exist. Run the macro 05_data_geometry. The selection of the orientation of this plane can be determined by the geometry of the three-dimensional zone containing the data points. such as: 'Strike: North 15 degrees East. These terms typically contain non-numeric information. however. 2. Dip: 40 degrees to the Northwest'. After reading the text below on the first form. After reading the text below on the second form. Page 86 of 137 . a plane defined as 'Striking N15E. coloured by Z. Dipping 40 degrees to the NW' could be defined in Surpac as either: Dip: -40 Dip direction: 285 or Dip: +40 Dip direction: 105 A solid model of the ore zone. If. If there is no particular orientation of greatest continuity. and two triangles. Prior to performing a geostatistical estimation of block model blocks. or primary variogram map. Geologists typically define the orientation of a planar feature. This plane will be used as the first. In this example. an anisotropic condition is said to exist.Variogram Maps Primary Variogram Map Primary Variogram Map 1. such as a vein-type ore zone with the terms 'strike' and 'dip'. 3. representing the two methods of defining the planes above will be displayed. click Apply. the data are contained in a vein-type ore zone. A variogram map is a tool which can be used to determine the presence of anisotropy for a given plane. The values of Dip: +40 Dip direction: 105 will be used in this example. The determination of isotropy or anisotropy is performed by comparing the ranges of variograms in a variety of orientations. which can be represented as a plane.

tcl. but note the name of the file and the directory structure (contained in the Address field) for future reference. c. b. in case you have not installed the reference manual for the software. This file is included with the data set for this tutorial.Variogram Maps Primary Variogram Map The data is displayed in graphics as shown: Definition of a plane to be used for a variogram map 4. You do not need to understand everything in that document immediately. Run macro 06_primary_variogram_map. The reference manual for variogram maps displayed A form displayed Take a minute to browse through the reference manual page. You will see: a. A variogram map displayed in three dimensions. Variogram map reference manual and form displayed Page 87 of 137 .

coloured by gamma(h) values. calculated at regular angular increments. Once you APPLY that form. the variogram modelling form will be presented.htm. The orientation of the major axis in this case has been selected as a 'pitch' or 'rake angle' of 15 degrees from the horizontal strike of the variogram map plane. it will take a few minutes to calculate the variogram map. 2) The secondary variogram map is perpendicular to the major axis. The interpretation of variogram data. There are two variogram maps: 1) The primary variogram map will contain the major axis. After reading the text below on the form. and the selection of the orientation of the direction of maximum continuity is done here to demonstrate the geostatistical functionality available. While it is processing. or read the reference manual page. it is important to understand a couple of basic concepts. minimise or close the window containing the reference manual page. you may click on this window to review the data. A variogram map is a set of experimental variograms.Variogram Maps Primary Variogram Map 5. When you have finished. Please refer to it for a detailed explanation of variogram maps. For the purpose of this demonstration. Page 88 of 137 . The primary variogram map. and has been included in this tutorial and opened by this macro. Many aspects of geostatistics are not an exact science. move the form to the side so that you can see the image in graphics. and contains the semimajor and minor axes.. After this form. This direction was selected based on an interpretation that the range of the variogram at this orientation was longer than the range of any other variogram on this plane. Note: Do NOT click Apply yet. The reference manual page for variogram maps is SSI_REFMAN:/default/stats/map. with Dip: +40 Dip direction: 105 has been displayed by this macro in the main graphics window.

The semimajor axis may or may not lie within the plane of the primary variogram map. the primary variogram map will contain the major axis of the anisotropy ellipsoid. The form below is displayed. 6. Choose Geostatistics > Variogram modelling to open the VARIOGRAM MODELLING window. Primary variogram map in 3D By definition. move the form back into view and click Apply.Variogram Maps Primary Variogram Map You will see the “final product” you will be creating. a primary variogram map. Choose Variogram map > New variogram map. 7. 8. Variogram map – Basic tab Page 89 of 137 . Once you have viewed the primary variogram map. in graphics.

Note that here you are using: Dip: +40 Dip direction: 105 Note: As was mentioned previously. the lag. the practicality of using a variogram map to visually determine anisotropy in a plane often rules out reducing the angular increment. Although this is a plausible argument in theory. and variogram report parameters are specified. maximum distance. 24 variograms will result in a 15 degree angular increment (360/24=15).Variogram Maps Primary Variogram Map In the top panel. the number of data pairs would be so small that very few. ID. For this data set. To some. You may find that you will need to try a few variations of this value to get one that is adequate. In the bottom panel. In the middle panel. The relationship between the angular increment and the spread angle should be considered. A spread of 30 degrees is used for this data set to ensure that sufficient samples are included.5 degree spread were used (half of the 15 degree angular increment between adjacent variograms). this same plane could be defined using a dip of -40 and a dip direction of 285. Page 90 of 137 . it may seem unreasonable to define a spread tolerance anything greater than half of the angular increment. string range. exactly as they are in variogram modelling. One item to consider – the maximum distance will be the radius of the variogram map. one could argue that the number of variograms should be reduced to minimise the “overlap” of the cones for adjacent variograms.) is defined using the same parameters as in variogram calculation. string file information (Location. The number of variograms selected will determine the angular increment. If the number of variograms was set to 36. In our example. you would get a 10 degree increment (360/36=10). if any reasonable variograms would result. etc. This is yet another example of how geostatistics is an inexact science. Experience with a data set will usually let you know what combination of parameters is acceptable. the plane containing the primary variogram map is defined. The spread and spread limit parameters are the same as in normal variogram modelling. because of the small number of pairs. Given that a spread of 30 degrees is used. It is up to you and the data set you are working with as to how you want to go about achieving both of these goals simultaneously. if a 7.

The variogram map currently displayed appears to be 'incomplete'.do not close the variogram modelling window or make any modifications to the data displayed. It has been coloured by gamma(h) values. -------------------------------------------------------------------------------------------------------------------- The primary variogram map. You may want to review and rotate the data there before proceeding with the next macro.the 'missing' parts of the variogram map are simply where no data pairs exist at that distance and orientation. or has 'missing' data. Click the Advanced tab. which are stored in the D1 field.tcl. must be run after this. You will see that the fields here are identical to those on the advanced tab of the variogram calculation form. click Apply. After reading the text below on the second form.dtm. click Apply. The next macro. DO NOT MODIFY ANY WINDOWS OR DATA DISPLAYED. and is displayed in the main graphics window. with Dip: +40 Dip direction: 105 is now displayed. Variogram map – Advanced tab 10. Page 91 of 137 . the variogram map has been saved as the file primary_variogram_map1. Additionally. _07_secondary_variogram_map. There is actually nothing wrong . 11. After you have completed viewing this. The orientation of the major axis has been selected as a 'pitch' or 'rake angle' of 15 degrees from the horizontal strike of the variogram map plane. and depends on the variogram modelling window presented as it is .Variogram Maps Primary Variogram Map 9.

For a horizontal seam deposit.Variogram Maps Primary Variogram Map This direction was selected based on an interpretation that the range of the variogram at this orientation was longer than the range of any other variogram on this plane. the major axis lies within the primary variogram map.Experimental variogram (SAVE EXP VG) has been used to save the file major_axis. the orientation of the primary variogram map would be horizontal. You will see the variogram map.Save . The function File . As described previously. as well as the variogram for the orientation displayed on the variogram map.vgm. interpretations of geostatistical data are performed here only to demonstrate the functionality of the software. Alternate interpretations may be deemed more appropriate.evg.Variogram model (SAVE VG MODEL) has been used to save the file major_axis.Save . Variogram map The purpose of the primary variogram map is to determine the orientation of the major axis. The example given here is based upon the premise that you. The function File . Page 92 of 137 . it is possible to review the variogram model by opening this file. by definition. as a geologist know that the orientation of the major axis will lie somewhere in this dipping plane for this dataset. When the experimental variogram data is displayed. It is possible to review the experimental variogram data by opening this file. Additionally. As previously stated.

You will probably need to change this value to fit your desired orientation. click and drag the red line to your selected orientation. Once you have an acceptable variogram. Once you have determined the orientation of the major axis. Notice that this is a free rotation. indicating the relative orientation of your selection. You will most likely see that throughout a range of lag values. c. but the orientation of the major axis is also open to interpretation and debate. Click Apply to enter the value. modify the model to fit that experimental variogram. not only is the subject of variogram modelling a non-scientific process. as a value between 0 and 180. 13. e. b. g. Repeat the previous two steps until you are satisfied that you have the orientation of the major axis. and you are not forced to select a precise orientation of any of the variogram directions. ask yourself and others who are familiar with the geology if the orientation appears correct. Modifying the lag distance for that orientation may help you get a better fit. 12. Use the Next direction (and/or Previous direction) function to rotate through all of the other orientations. and others which will be consistently low. This can be useful if you have found two adjacent orientations which are equally valid – you can set the direction of maximum continuity to midway between these two directions. As you can see. After you have ascertained a major axis. If another orientation appears to have a longer range and a lower variance. then release. The lag slider and the “Next direction” functions can help you. d. The major axis should be that variogram which has the lowest variance for the longest distance. Choose Variogram map > Select direction of maximum continuity. note that the orientation of 15 degrees above the horizontal (on the left) will consistently display colours on the low end of the variance values. Page 93 of 137 . and watch the colours on the variogram map change. Now use the lag slider to improve the quality of the experimental variogram for that direction. move the lag slider back and forth. Once you have an idea of what may appear to be the orientation of the longest range. In the variogram map viewport. Direction of maximum continuity 14. In other words. use the Next direction function to rotate the black line on the variogram map to that direction. Use the lag slider to alert you to areas of high and low variance on the variogram map. there will be areas on the variogram map which will be consistently high. h. Using the example given above. create a variogram model for that orientation. as represented by the legend to the right. as described here: a. f. A form will display. you must inform the software of your selection.Variogram Maps Primary Variogram Map The idea is to select the variogram orientation which has the longest range for a given sill value.

Variogram Maps Secondary Variogram Map Secondary Variogram Map 1. Run the macro 07_secondary_variogram_map. A secondary variogram map will now be created. perpendicular to the semi-major axis. 2. Once you Apply this form. After reading the text below on the first form. The orientation of the minor axis is automatically calculated based on the orientation of the semi-major axis . and the selected orientations of all axes. Click on the main window to review this data. are displayed in graphics: Primary and Secondary variogram maps Page 94 of 137 . The selected direction of maximum continuity in the secondary variogram map is the semimajor axis. A DTM of the secondary variogram map has been displayed in graphics. click Apply. The minor axis will lie in the plane of the secondary variogram map. it will take a few minutes to calculate the secondary variogram map. in this case. While it is processing.tcl. you can review the secondary variogram map in the main window or read the reference manual page on variogram maps. The secondary variogram map. Note that the semi-major axis. is not contained in the plane of the primary variogram map.no user input is required. although it could have been.

must be run after this. In this case. The orientation of the minor axis is automatically calculated based on the orientation of the semi-major axis . the variogram map has been saved as the file secondary_variogram_map1. click Apply.5 Additionally.no user input is required. This orientation is the semi-major axis. -------------------------------------------------------------------------------------------------------------------- The secondary variogram map has been calculated. The minor axis is also in the plane of the secondary variogram map. and is perpendicular to the semi-major axis.42 Dip direction: 26.do not close the variogram modelling window or make any modifications to the data displayed. It has been coloured by gamma(h) values. The next macro.42 Dip direction: 206.DTM. The secondary variogram map has been created perpendicular to the primary variogram map (Dip: +40 Dip direction: 105) and perpendicular to the major axis. the semi-major axis has been selected at an orientation slightly different to the dip of the primary variogram map. 'missing' data in any variogram map is a result of lack of data pairs at certain distances and orientations. DO NOT MODIFY ANY WINDOWS OR DATA DISPLAYED. The secondary variogram map has the orientation of: Dip: -80. and is displayed in the main graphics window. which are stored in the D1 field.5 which is the same as: Dip: 80. If the orientation of the semi-major axis was significantly different than the primary variogram map. it may be worth recalculating the primary variogram map at an orientation which encompasses both the major and semi-major axes as determined here. Page 95 of 137 .tcl. _08_anisotropy_ellipsoid. and depends on the variogram modelling window presented as it is . Again.Variogram Maps Secondary Variogram Map 3. and a direction of maximum continuity has been identified. You may want to review and rotate the data there before proceeding with the next macro. Notice that the same steps used to select the direction of maximum continuity for the primary variogram map have been used to select the direction of maximum continuity for the secondary variogram map. After reading the text below on the next form.

To determine these ratios. the major. After reading the text below on the first form. For block model estimation. and the experimental variogram data for the minor axis is presented. If the continuity of the data is the same in all directions. However. As mentioned previously. the semi-major axis is the second-largest measure of continuity. and the experimental variogram data for the semi-major axis is presented 3) The range of the variogram is graphically modified to match the data for the semi-major axis (the nugget and sill values of the variogram model are held constant). in this demonstration. the range of the variogram is graphically modified to match the data for the minor axis (the nugget and sill values of the variogram model are held constant). In other words. 7) The major to minor anisotropy ratio is calculated as: The length of the range of the variogram model for the major axis / The length of the range of the variogram model for the minor axis The two anisotropy ratios are calculated and displayed in the upper right viewport. an 'anisotropic' condition exists. and minor axes would all have the same continuity. These ratios are calculated by comparing the length of the ranges of the variograms along each axis. The lag distance may be modified to improve the quality of the variogram. By definition. an 'isotropic' condition is said to exist. as well as the strength of the continuity relative to the other axes is described by an 'anisotropy ellipsoid'. The orientation of the three axes. the major axis is the orientation which represents the greatest continuity.Variogram Maps Anisotropy Ellipsoid Parameters Anisotropy Ellipsoid Parameters 1. and the minor axis is the smallest. as there is variable continuity. Page 96 of 137 . semi-major. Run the macro 08_anisotropy_ellipsoid. 1) The function 'Extract variograms along axes' (VG SHOW AXES VARIOGRAMS) is performed. 6) As in step 3. it is necessary to know the strength of the major axis as compared to the strength of the other two axes.tcl. 2. and the experimental variogram data and the variogram model for the major axis is displayed. 4) The major to semi-major anisotropy ratio is calculated as: The length of the range of the variogram model for the major axis / The length of the range of the variogram model for the semi-major axis 5) The 'Next variogram' (VG NEXT) function is performed again. The lag distance may be modified to improve the quality of the variogram. The strength of the continuity along the major axis is compared to the continuity of the semimajor and minor axes by the following two terms: The major to semi-major anisotropy ratio and the major to minor anisotropy ratio. 2) The 'Next variogram' (VG NEXT) function is performed. click Apply.

After this. click Apply. Page 97 of 137 . and the experimental variogram data for the major axis is displayed. the macro will modify the range of the variogram (normally this is done graphically by the user) to match the data for the semi-major axis (the nugget and sill values of the variogram model for the major axis are held constant).Open Variogram model (RECALL VG MODEL) has been used to display the major axis variogram model. along with the major axis variogram model. File . After reading the text below on the next form. The 'Extract variograms along axes' (VG SHOW AXES VARIOGRAMS) function has been performed. and the experimental variogram data for the semi-major axis will be presented.Variogram Maps Anisotropy Ellipsoid Parameters 3. Once this is done. the major to semi-major anisotropy ratio will be calculated as: The length of the range of the variogram model for the major axis / The length of the range of the variogram model for the semi-major axis This anisotropy ratio will be displayed in the upper right viewport. The 'Next variogram' (VG NEXT) function will be performed.

5. The range of the variogram has been modified to match the data for the semi-major axis. click and drag it out of the way so that you can view the variogram modelled for the semi-major axis.Variogram Maps Anisotropy Ellipsoid Parameters The secondary variogram map. When the next form is displayed. 4. After moving the form back and reading the text below on the form. Note that the nugget and sill values of the variogram model are held constant. The ranges of the variogram models for the major and semi-major axes are used to calculate the major to semi-major anisotropy ratio: The length of the range of the variogram model for the major axis / The length of the range of the variogram model for the semi-major axis Page 98 of 137 . click Apply. and the selected orientations of all axes are displayed.

6. the 'Next variogram' (VG NEXT) function will be performed. Page 99 of 137 . along with the current variogram model. the major to semi-major anisotropy ratio is: 168 / 99 = 1. and the experimental variogram data for the minor axis will be presented.7 This anisotropy ratio is displayed in the upper right viewport. the length of the ranges of the variogram models are: Major axis: 168 Semi-major axis: 99 Therefore. click and drag it out of the way so that you can view the variogram modelled for the minor axis. Following this. The major / minor anisotropy ratio will be calculated in the same manner. When the next form is displayed.Variogram Maps Anisotropy Ellipsoid Parameters In this example.

After moving the form back and reading the text below on the form. Page 100 of 137 . The 'Next variogram' (VG NEXT) function has been performed. click Apply. Data such as this presents a challenge for interpretation. in terms of Surpac's rotation convention. the major to minor anisotropy ratio is: 168 / 33 = 5. The lag distance has been changed to improve the quality of the variogram.1 This anisotropy ratio is displayed in the upper right viewport. The function 'Create anisotropy ellipsoid report' will be used to do this. and the experimental variogram data for the minor axis has been presented. The range of the variogram has been modified to match the data for the minor axis. The next step is to determine the orientation of the anisotropy ellipsoid. Note that the requirement of maintaining the same nugget and sill values of the variogram model does not allow a very good fit of the variogram in this case. The ranges of the variogram models for the major and minor axes are used to calculate the major to minor anisotropy ratio: The length of the range of the variogram model for the major axis / The length of the range of the variogram model for the minor axis In this example.Variogram Maps Anisotropy Ellipsoid Parameters 7. along with the semi-major axis variogram model. the length of the ranges of the variogram models are: Major axis: 168 Minor axis: 33 Therefore.

7 Major to minor anisotropy ratio: 5.4 4.0 Major to semi-major anisotropy ratio: 1.Variogram Maps Anisotropy Ellipsoid Parameters 8. click Apply.tcl.6 Plunge: -9.do not close the variogram modelling window or make any modifications to the data displayed.4 A (range) : 168. DO NOT MODIFY ANY WINDOWS OR DATA DISPLAYED. After reading the text below on the next form.1 Page 101 of 137 . -------------------------------------------------------------------------------------------------------------------- Now that you have obtained all of the information below. and depends on the variogram modelling window presented as it is . The next macro. _09_create_block_model.6 Dip: -45 Major axis variogram parameters: Co (nugget): C (sill) : 5. you are ready to proceed with block model estimation: Anisotropy ellipsoid orientation (Surpac convention): Bearing: 206. must be run after this.

Variogram Maps Anisotropy Ellipsoid Parameters 9. and you will see the ellipsoid displayed: Page 102 of 137 . Click on the main Surpac window.

Click and drag the red line on the variogram map until it is aligned with the orientation of the major axis. Modifying the lag distance for that orientation may help you get a better fit.VGM for this orientation. Choose Variogram map > New variogram map. 13. but can be helpful in a future step. 4. Choose Variogram map > Select direction of maximum continuity. which is the major axis. The direction of maximum continuity (the red line) will display as the intersection of the primary and secondary variogram maps. You will notice that very small lag values are usually not useful. you may wish to select Display/Hide variance to show the data variance (often used as the total sill). Choose Geostatistics > Variogram modelling to open the variogram window.Variogram Maps Steps for Using Variogram Maps to Create Anisotropy Ellipsoid Parameters Steps for Using Variogram Maps to Create Anisotropy Ellipsoid Parameters In summary. and others which will be consistently low. Create a variogram model for this orientation. under the Display menu) may help you to view this variogram. you may wish to select Display/Hide variance to show the data variance (often used as the total sill). under the Display menu) may help you to view this variogram. Once you have an idea of what may appear to be the orientation of the longest range. The orientation of this line should be relatively close to what will become the semi-major axis. Now use the lag slider to improve the quality of the experimental variogram for that direction. 15. which will become the semi-major axis. as represented by the legend to the right. Use the lag slider to alert you to areas of high and low variance on the variogram map. there will be areas on the variogram map which will be consistently high. You may also find that Set axis limits function (again. Page 103 of 137 . Choose File > Save > Variogram model to save a *. 10. This is an optional step. 9. 17. 16. 5. here are the complete set of steps to obtain all of the anisotropy ellipsoid parameters: 1. note that the orientation of 15 degrees above the horizontal (on the left) will consistently display colours on the low end of the variance values. From the Display menu. From the Display menu. but may help you to display the orientation of the primary variogram map in three dimensions in graphics. This is an optional step. Create a variogram model for that orientation. Enter the variogram map parameters and click Apply. Repeat the previous two steps until you are satisfied that you have the orientation of the major axis. If another orientation appears to have a longer range and a lower variance than your current model. and watch the colours on the variogram map change. 3. You will most likely see that throughout a range of lag values. 11. use the Next direction function (under the Display menu) to rotate the black line on the variogram map to that direction. Using the example given above. move the lag slider back and forth. 2. Choose Variogram map > Save DTM. modify the model to fit that experimental variogram. 7. Now use the lag slider to improve the quality of the experimental variogram for that direction. 6. Use the Next direction (and/or Previous direction) function to rotate through all of the other orientations. 12. 14. You may also find that Set axis limits function (again. In other words. Choose Variogram map > Secondary Variogram map. The major axis should be that variogram which has the lowest variance for the longest distance. Use the Next direction (and/or Previous direction) function to rotate the black line on the variogram map to that direction. 8.

Note: Page 104 of 137 . Choose Variogram map > Change anisotropy. Choose Variogram map > Save DTM. Use the Next direction (and/or Previous direction) function to rotate through all of the other orientations. You may need to use the lag slider to improve the quality of the variogram. the range of the semi-major axis must be equal to or longer than the range of the minor axis for a given sill. you may choose to continue. By definition. Repeat the previous two steps until you are satisfied that you have the orientation of the major axis. then you need to start again from step 1. If the variogram representing the semi-major axis is to the right of the model for the major axis. then you need to start again from step 14. modify the model to fit that experimental variogram. then Variogram model). If you did not previously save a variogram model. Modifying the lag distance for that orientation may help you get a better fit. 31. and one of all variograms together. but you do have good quality variograms for the major and semi-major axes. The experimental variogram for the major axis will be displayed in red. or to the left of the variogram model for the semi-major axis. 28. use the Next direction function to advance to the next variogram. 30. By definition. the range of the major axis must be equal to or longer than the range of the semi-major axis for a given sill. create a variogram to fit the major axis (from the Variogram > Model). The semi-major axis should be that variogram which has the lowest variance for the longest distance. the major axis. Use the Next direction function (and/or Previous direction) to advance through the four views of the experimental variograms – one each of the major. 19. 27. If the variogram representing the minor axis is to the right of the model for the semi-major axis. This is an optional step. Modify the lag if required to improve the quality of the experimental variogram. Click and drag the sill/range marker to the left until the variogram model matches the experimental variogram for the semi-major axis. Ensure that the variogram for the semi-major axis is either the same as. 20. and the minor axes. Ensure that the variogram model fits the experimental variogram. The current minor axis is now a more likely candidate for the orientation of the semi-major axis. Once you have assured yourself that your variograms are acceptable. and determine the ratio for the minor axis based on other factors. the semi-major. the semi-major axis. Ensure that the variogram for the minor axis is either the same as. If another orientation appears to have a longer range and a lower variance than your current model. The current semi-major axis is now a more likely candidate for the orientation of the major axis. 22. 21. display it now (from the File > Open. Again use the Next direction function to advance to the next variogram. but may help you to display the orientation of the secondary variogram map in three dimensions in graphics. 24. If you cannot get a visually acceptable minor variogram. Stop when you get to the view of all three variograms. Choose Variogram map > Select direction of maximum continuity. 25. 26. using the orientation of the current semi-major axis within the primary variogram map. You may need to use the lag slider to improve the quality of the variogram. such as geometry. or to the left of the variogram model for the major axis. 29. 23.Variogram Maps Steps for Using Variogram Maps to Create Anisotropy Ellipsoid Parameters 18. Choose Variogram map > Extract variograms along axes. You will not be able to modify either the nugget or the sill – only the range is changed to calculate the anisotropy ratio. Note: It is often difficult or impossible to interpret the experimental variogram for the minor direction. 32. Click and drag the red line on the variogram map until it is aligned with the orientation of maximum continuity. If you saved a *.VGM (variogram model) for the major axis.

Use the Next direction function to advance to the next variogram. Click and drag the sill/range marker to the left until the variogram model matches the experimental variogram for the minor axis. 37. the minor axis.Variogram Maps Steps for Using Variogram Maps to Create Anisotropy Ellipsoid Parameters 33. 36. 35. you may choose modify the range until the ratio for the minor axis is equal to some value you have chosen based on other factors. Choose File > Close to exit the variogram modelling window. Choose Variogram map > Create anisotropy ellipsoid report. but you do have good quality variograms for the major and semi-major axes. 38. If you cannot get a visually acceptable minor variogram. Choose Variogram map > Ellipsoid visualiser. You can view or save the ellipsoid. Modify the lag if required to improve the quality of the experimental variogram. Page 105 of 137 . as well as the major/semi-major and major/minor anisotropy ratios. Summary You should now understand the following terms: • • • Variogram map Primary variogram map Secondary variogram map You should also know how to create and use variogram maps to determine the parameters which define an anisotropy ellipsoid with Surpac. such as geometry. Note: It is often difficult or impossible to interpret the experimental variogram for the minor direction. This report will contain values for the orientation of the anisotropy ellipsoid. 34. Choose Variogram map > Change anisotropy.

Isotropic vs anisotropic inverse distance estimation. you should understand the following concepts: • • • • • • Surpac menubars Surpac string files Surpac block models isotropy and anisotropy anisotropy ellipsoid the parameters which define an anisotropy ellipsoid Page 106 of 137 . or a set of points in space which contain estimated values. The impact of power on inverse distance estimation. One of the methods for estimating values at points in a model is known as inverse distance estimation. described in this chapter through the following: 1. Steps to performing inverse distance estimation.Inverse Distance Estimation Steps for Using Variogram Maps to Create Anisotropy Ellipsoid Parameters Inverse Distance Estimation Overview An important end product of a geostatistical evaluation is a “model”. 3. Requirements Prior to proceeding with this chapter. 2.

Page 107 of 137 .Inverse Distance Estimation Isotropic vs Anisotropic Inverse Distance Estimation Isotropic vs Anisotropic Inverse Distance Estimation When using inverse distance weighting to estimate values in a block model. anisotropy can be used in any estimation technique. 2. the amount and direction of anisotropy can have a significant impact on the end result. Click in graphics after each model is displayed. but different amounts and orientations of anisotropy were used.tcl to see how these blocks are displayed in Surpac. you can see all values on the forms by ticking “Slow motion playback”. See the previous chapter entitled “Anisotropy” for more information. If you use the Macro playback button. the four models shown below were created from the same data set. Although anisotropy is one manner in which inverse distance weighting can impact the end result. Run the macro anisotropy. For example. No Anisotropy (Isotropic) 2:1 Anisotropy Azimuth 45 2:1 Anisotropy Azimuth 135 5:1 Anisotropy Azimuth 135 1.

choose Display > Clear model colours.Inverse Distance Estimation Steps to Performing Inverse Distance Estimation Steps to Performing Inverse Distance Estimation 1. 8. Page 108 of 137 .mdl. 6. 3. and then click Apply. Choose Display > Point > Markers. 7. From the Block Model menubar. From the Block Model menubar. 4. Enter the information as shown. and then click Apply. 10. Choose Display > Point > Attributes. Zoom out and you will see the points as shown: This is the data set you will use for estimating blocks in the model. Enter the information as shown. 9. choose Display > Display block model.str. Open the model 20x20x1. 5. Display the Block Model menubar. Open 4points1. 2.

Inverse Distance Estimation Steps to Performing Inverse Distance Estimation 11. and then click Apply. and then click Apply on each form. Enter the information as shown. You should now see the model and the data displayed as shown: 12. Enter the information as shown. Choose Estimation > Inverse distance. Page 109 of 137 .

Choose Display > Colour model by attribute. Page 110 of 137 .Inverse Distance Estimation Steps to Performing Inverse Distance Estimation 13.

The model. click Apply. click Scan. Once you have set up the parameters on the form. On the Block Colours form.Inverse Distance Estimation Steps to Performing Inverse Distance Estimation 14. select isotropic as the attribute to color by. 15.70. enter values for Range for colour selection of 30. as estimated by inverse distance squared using an isotropic search is displayed as shown: Page 111 of 137 .5 and then click Refresh.

Choose Attributes > View attributes for one block. Notice that the attributes distance. and num_samp have been added to the model: Page 112 of 137 . avg_distance.Inverse Distance Estimation Steps to Performing Inverse Distance Estimation 16.

Inverse Distance Estimation The Impact of Inverse Distance Power The Impact of Inverse Distance Power 1. 4. Choose Estimation > Inverse distance. Enter the data as in the previous example. as before. 3. 2. except on the last form enter an Inverse Distance Power of 3 before clicking Apply. Power of 2 Power of 3 Power of 4 Power of 10 The effect of power on Inverse distance estimation Page 113 of 137 . but using inverse distance powers of 4 and 10. Repeat the process of estimating and displaying the attribute isotropic. Display the model coloured by the isotropic attribute.

you should now be aware of the impact of power on inverse distance estimation. more research. Page 114 of 137 .Inverse Distance Estimation The Impact of Inverse Distance Power Summary You should understand how anisotropy ratios and orientation of the anisotropy ellipsoid impacts the weight used for samples in estimating a value at a block centroid. Understanding how each of the many parameters associated with an individual estimation method is no simple task. It usually will take some time. Also. and/or experience with several data sets to grasp the concepts presented here.

The impact of the nugget effect. One of the methods for estimating values at points in a model is known as ordinary kriging. The impact of block descretisation. 3. Requirements Prior to proceeding with this chapter.Ordinary Kriging The Impact of Inverse Distance Power Ordinary Kriging Overview An important end product of a geostatistical evaluation is a “model”. The impact of the range. or a set of points in space which contain estimated values. described in this chapter through the following: 1. 2. you should understand the following concepts: • • • • • • Surpac menubars Surpac string files Surpac block models isotropy and anisotropy anisotropy ellipsoid the parameters which define an anisotropy ellipsoid Page 115 of 137 .

tcl.Ordinary Kriging Impact of the Nugget Effect Impact of the Nugget Effect When using ordinary kriging to estimate values in a block model. run the macro ordinary_kriging_nugget. the relative percentage of the nugget can have a significant impact on the end result. 5. Click in graphics to view each result. To see these blocks displayed in Surpac. Page 116 of 137 . zero nugget 25% nugget 50% nugget 100% nugget Impact of the nugget 4.

tcl. run the macro ordinary_kriging_range. Click in graphics to view each result.5x sample spacing) Range = 2.5 (0. To see these blocks displayed in Surpac.25x sample spacing) Impact of the range relative to sample spacing (Sample spacing = 10) 1. Page 117 of 137 . Range = 20 (2x sample spacing) Range = 10 (1x sample spacing) Range = 5 (0. 2. the range of the variogram relative to sample spacing can have a significant impact on the end result.Ordinary Kriging Impact of the Range Impact of the Range When using ordinary kriging to estimate values in a block model.

7. From the Block Model menubar. Display the data as markers.str. 4. 8. Open the file 4points1. Enter the information as shown. choose Display > Clear model colours. and then click Apply. You should now see the model and the data displayed as shown: Page 118 of 137 . From the Block Model menubar.mdl.mdl. 9. Display the Block Model menubar. choose Display > Display block model.Ordinary Kriging Impact of the Range You will now go through each step in the process of ordinary kriging estimation. This is the data set you will use for estimating blocks in the model. using an isotropic estimation with the model 20x20x1. 3. 6. Ensure that you have opened the model 20x20x1. and display the D1 field. 5.

Choose Estimation > Ordinary kriging. and then click Apply on each form.Ordinary Kriging Impact of the Range 10. 11. Page 119 of 137 . Enter the information as shown.

Ordinary Kriging Impact of the Range Page 120 of 137 .

Choose Display > Colour model by attribute. Click Apply on the two following forms to overwrite the *. click Scan. avg_distance.5 and then click Refresh.70. as estimated by ordinary kriging using a zero nugget effect is displayed: 16. enter values for Range for colour selection of 30. 14. The model. and krig_var have been added to the model: Page 121 of 137 . Once you have set up the parameters on the form.NOT file and the *. 15.MDL file. On the Block Colours form. click Apply.Ordinary Kriging Impact of the Range 12. num_samp. Choose Attributes > View attributes for one block. select isotropic as the attribute to color by. Notice that attributes distance. 13.

Page 122 of 137 . more research. range. Understanding how each of the many parameters associated with an individual estimation method is no simple task. and descretisation points impact the outcome of an ordinary kriging estimation. 18. but using descretisation of 2x2x2 and 3x3x3. 1x1x1 2x2x1 2x2x2 3x3x3 The effect of descretisation on ordinary kriging estimation Summary You should understand how nugget. except on the last form. Enter the data as above. 19. as before. and/or experience with several data sets to grasp the concepts presented here. Repeat the process of estimating and displaying the attribute isotropic. enter descretisation parameters of 2x2x1 before clicking Apply.Ordinary Kriging Impact of the Range 17. 20. Choose Estimation > Ordinary kriging. It usually will take some time. Display the model coloured by the isotropic attribute.

Block site selection. The analysis of the block size is described in this chapter through the following: 1. Debug output from ordinary kriging. may be based on other factors. Requirements Prior to proceeding with this chapter. The determination of the spacing between these points. or a set of points in space which contain estimated values. Using kriging efficiency and conditional bias slope. The points are representative of the centroid of a block of material. however. 3. The ultimate selection of the block size. 2.Block Size Analysis Impact of the Range Block Size Analysis Overview An important end product of a geostatistical evaluation is a “model”. such as minimum mining width. By evaluating two parameters derived from ordinary kriging. or the block size is often a critical factor in a geostatistical estimation. you should understand the following concepts: • • • • • • Surpac menubars Surpac string files Surpac block models isotropy and anisotropy anisotropy ellipsoid ordinary kriging Page 123 of 137 . you can determine the optimal block size.

000 0. creates a block model performs ordinary kriging within a constraint to estimate a value for a single block uses debug output in the report file This output file. block_size. Near the bottom of the file. block variance – kriging variance Kriging Efficiency = block variance Block variance – kriging variance + | lagrange multiplier | Conditional Bias Slope = Block variance – kriging variance + 2x | lagrange multiplier | Page 124 of 137 . 1. you will see the following parameters: Estimated grade: Kriging variance: Twice std. This macro: a.Block Size Analysis Debug Output from Ordinary Kriging Debug Output from Ordinary Kriging When using ordinary kriging to estimate values in a block model.907 0.930 The values of kriging efficiency and conditional bias slope are used in analysing block size.452 0. b.not will be displayed.053 0. Run the macro kriging_debug. the debug output will contain two parameters which will be used in the analysis of block size.: Block variance: Kriging efficiency: Slope of regression: Lagrange multiplier: Conditional bias slope: 4.112 0. dev.250 1. c.724 0.tcl.

907 0. To do this in Surpac. Y. you would perform the following functions for a set of block sizes: 1. 4.930 Page 125 of 137 . 3. 2.250 1. this is impossible.: Block variance: Kriging efficiency: Slope of regression: Lagrange multiplier: Conditional bias slope: 4. Z<115 debug output ticked output file: block_size. Y. 5. In practice. and Z block dimensions create a block model using an origin such that the coordinates of the centroid of the first block in the model are the same as the coordinates in step 1 performs ordinary kriging within a constraint to estimate a value for the first block ensure “debug output” option is used in the report file note the values for kriging efficiency and conditional bias slope in the debug file Here is the data from the kriging_debug.724 0. and Z coordinates of a location where you wish to perform the analysis select the X.112 0.000 0. 6. select the X. but what can be done is to compare values of these two parameters for a variety of block sizes. dev.Block Size Analysis Using Kriging Efficiency and Conditional Bias Slope Using Kriging Efficiency and Conditional Bias Slope Ideally. both the kriging efficiency and the conditional bias slope should have a value of 1. X<1665.not contained: Estimated grade: Kriging variance: Twice std.053 0.tcl example: • • • • • • coordinates to perform the analysis: Y=7340 X=1660 Z=110 block dimensions: Y=10 X=10 Z=10 block model origin: Y=7335 X=1655 Z=105 ordinary kriging within constraint: Y< 7345.00.452 0.

.5.5 X=1657. Z<111 Page 126 of 137 . Z<112. Z<111 • • • block dimensions: Y=2 X=2 Z=2 block model origin: Y=7339 X=1659 Z=109 ordinary kriging within constraint: Y< 7341.5 ordinary kriging within constraint: Y< 7345.Block Size Analysis Using Kriging Efficiency and Conditional Bias Slope Given these same data in step 1.5 Z=109 ordinary kriging within constraint: Y< 7342. X<1661. In each example. coordinates to perform the analysis: Y=7340 X=1660 Z=110 • • • block dimensions: Y=10 X=5 Z=5 block model origin: Y=7335 X=1657.. the coordinates of the centroid will be the same. X<1662.5 • • • block dimensions: Y=5 X=5 Z=2 block model origin: Y=7337. here are some examples for different block sizes. X<1662..5.5 Z=107.

you would want to choose a location which is in the middle of the data. search radius. but are related to block size. sample locations. and the procedure for optimising a set of block sizes. each representative of different types of data clustering.tcl example: However. Summary You should understand how to obtain values for kriging efficiency and conditional bias slope. 2. if there were one or more clusters within the data. 5. as the kriging efficiency and conditional bias slope are determined by block size and sample locations near the block. 3. as is the case with the string file gold_cut17.Block Size Analysis Block Site Selection Block Site Selection If your data is fairly regular. 1. Here is the block site location and string file data used in the kriging_debug. Page 127 of 137 . you can use the procedure described here to compare other factors.str. and variogram parameters. area representative of average sample spacing within or near cluster #1 within or near cluster #2 within or near cluster #3 within or near cluster … Kriging efficiency and conditional bias slope are not dependent upon sample values. 4. and discretisation points. such as maximum number of samples. Once you have determined an optimal block size. you should perform the study at each of several locations.

There are several means of performing this: 1. Comparing cross-sectional data with model values Grade-tonnage curves from block model reports Basic statistics of model values Trend analysis Requirements Prior to proceeding with this chapter.Model Validation Block Site Selection Model Validation Overview An important step in a geostatistical evaluation is to validate the model after it has been created. 2. 3. you should understand the following concepts: • • • • • • Surpac menubars Surpac string files Surpac block models isotropy and anisotropy anisotropy ellipsoid ordinary kriging Page 128 of 137 . 4.

as well as kriged values of blocks inside the solid: Page 129 of 137 . raw and composited drillhole data. This macro will display a cross-section of a 3D solid. you want to ensure that the values in the model appear to be correct. 2.tcl. and the calculated partial percentage of blocks inside the solid: Basically. Run the macro _15a_display_partial_percent.Model Validation Comparing Cross-sectional Data with a Model Comparing Cross-sectional Data with a Model One method of validating a model is to view cross-sections of it compared to other data.00. This macro will display a cross-section of a 3D solid.tcl. This does appear to be the case: blocks entirely within the model have a partial percent value of 1. 1.00 and blocks partially within the model have a partial percentage value between 0 and 1. Run the macro _15b_display_blocks&composites.

Model Validation Comparing Cross-sectional Data with a Model Again. Page 130 of 137 . you want to check that block values and model values appear to be correct.

XLS file is displayed at the end with a graph of the grade-tonnage curve. A predefined *. 2. Click Apply on each of the forms displayed: Page 131 of 137 . This macro will perform block model reporting to create a *.tcl. 1.CSV file containing grade and tonnes.Model Validation Grade-Tonnage Curves Grade-Tonnage Curves Another means of validating a model is to report tonnes and grade and construct a grade-tonnage curve. Run the macro _16a_bm_report.

Model Validation Grade-Tonnage Curves The file grade_tonnage.xls has been prepared with a graph of the output data: Page 132 of 137 .

Ensure the information as shown. and then click Apply. Run the macro _12_bm_basic_stats. This macro will display basic statistics on three block model parameters. 1. 2. You will see a histogram of the data displayed as shown. Page 133 of 137 .Model Validation Basic Statistics of Model Values Basic Statistics of Model Values Basic statistics of the block model values is yet another way to validate the output from the model.tcl.

2. Run the macro _16b_export_centroids. 1. Click Apply on any forms: Page 134 of 137 .Model Validation Trend Analysis Trend Analysis Basic statistics of the block model values is yet another way to validate the output from the model.tcl.

you will see the data in graphics: 3. This macro performs the function Geostatistics > Trend analysis. The main tab contains the names of the file(s) to be analysed: Page 135 of 137 . Notice that there are two tabs on the form which is displayed.Model Validation Trend Analysis When the data has been exported.tcl. Run the macro _16c_bm_vs_composites.

xls contains the data and a graph: Page 136 of 137 . you want to compare the downhole composite values against the block centroids for a range of northings. Click Apply on the form when you have reviewed it. In this case.Model Validation Trend Analysis The “Co-ordinate Constraints” tab allows you to specify a range of coordinate values. 4. The previously prepared file bm_vs_composites.

3. Comparing cross-sectional data with model values Grade-tonnage curves from block model reports Basic statistics of model values Trend analysis Page 137 of 137 .Model Validation Trend Analysis Trend analysis can be used to identify regions where block model estimations may be different to the composited data. using the following techniques: 1. 4. Summary You should understand how to validate a model after it has been created. 2.

- Standard Operation Procedure
- ...9366213^B^26-Mar-2015^0
- ob dump
- IJEMHSJune2014!1!6 Somesh
- Rock Abrasivity
- 12 Tunnels in Weak Rock
- Dump Failure
- geostatistical ore reserve
- 07 Somanchi A774w
- 2.1 Work Study
- HTML
- Answer 2014
- 187476
- 3 Pager Application Form - HZL
- MOEF 2006 Notification
- Examination Schedule End Sem Modified
- Cv Govindvilluri
- 09 Direct Shear Test
- ventilation control.ppt
- QAThYDsrgzUC
- Solids
- Drill and Blast
- Geological Database
- Introduction

Sign up to vote on this title

UsefulNot usefulgeostatistics mining

geostatistics mining

- Plotting
- Block-Modelling surpac 6.1.pdf
- Surpac Geological Database Tutorial 2
- 101026847 Surpac Extracting and Processing Data
- Drill and Blast
- Solids Surpac
- Surpac DTM Surface Tutorial
- geostatistics
- Underground Mine Design
- Geological Database
- Surpac Pit_optimiser Tutorial
- Handbook Surpac
- Pit Design
- Surpac Introduction Tutorial
- Surpac Geological Database_63
- geostatistics
- Tutorial Vulcan by Venkat
- 16611356 Training Manual Block Modelling
- Surpac Vision Geostatistics
- Dtm Surfaces
- Basic Geostatistics
- Surpac Quick Start Tutorial 2
- Surpac Interpolator
- Whittle Introduction Tutorial
- Training Manual Block Modelling 2
- Surpac Open Pit Mine Design[1] 0
- Getting Started
- Manual Mine Sight Geoestadistica
- Surpac Solids Tutorial
- Integrated Geology, Resource Modelling, Mine Planning and Production
- Geo Statistics

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