Kriging

What is Kriging?
IDW and Spline, discussed earlier, are referred to as deterministic interpolation methods because they are directly based on the surrounding measured values or on specified mathematical formulas that determine the smoothness of the resulting surface. A second family of interpolation methods consists of geostatistical methods, such as Kriging, that are based on statistical models that include autocorrelationthe statistical relationship among the measured points. Because of this, not only do these techniques have the capability of producing a prediction surface, but they can also provide some measure of the certainty or accuracy of the predictions. Kriging is similar to IDW in that it weights the surrounding measured values to derive a prediction for an unmeasured location. The general formula for both interpolators is formed as a weighted sum of the data: arrangement in the weights, the spatial autocorrelation must be quantified. Thus, in Ordinary Kriging, the weight, li, depends on a fitted model to the measured points, the distance to the prediction location, and the spatial relationships among the measured values around the prediction location. To make a prediction with Kriging, two tasks are necessary: (1) to uncover the dependency rules and (2) to make the predictions. To realize these two tasks, Kriging goes through a two-step process: (1) variograms and covariance functions are created to estimate the statistical dependencecalled spatial autocorrelation values, which depends on the model of autocorrelationfitting a modeland (2) prediction of unknown values are made. It is because of these two distinct tasks that it has been said that Kriging uses the data twice: the first time to estimate the spatial autocorrelation of the data and the second time to make the predictions.

(s ) = Z (s ) Z 0 i i
i =1

Variography
Fitting a model, or spatial modeling, is also known as structural analysis, or variography. In spatial modeling of the structure of the measured points, we begin with a graph of the empirical semivariogram, computed as: Semivariogram(distance h) = 0.5 * average[ (value at location i value at location j)2] for all pairs of locations separated by distance h. The formula involves calculating the difference squared between the values of the paired locations. The diagram that follows shows the pairing of one pointthe red pointwith all other measured locations. This process continues for each measured point.

where Z(si) is the measured value at the ith location; li is an unknown weight for the measured value at the ith location; s0 is the prediction location; N is the number of measured values. In IDW, the weight, li, depends solely on the distance to the prediction location. However, in Kriging, the weights are based not only on the distance between the measured points and the prediction location, but also on the overall spatial arrangement among the measured points and their values. To use the spatial

PERFORMING SPATIAL ANALYSIS

141

Kriging interpolation
There are two kriging methods: Ordinary and Universal. Ordinary kriging is the most general and widely used of the kriging methods and is the default. It assumes the constant mean is unknown. Universal kriging should only be used when you know there is a trend in your data and you can give a scientific justification to describe it. By using a variable search radius, you can specify the number of points to use in calculating the value of the interpolated cell. This makes the search radius variable for each interpolated cell, depending on how far it has to stretch to reach the specified number of input points. Specifying a maximum distance limits the potential size of the radius of the circle. If the number of points is not reached before the maximum distance of the radius is reached, fewer points will be used in the calculation of the interpolated cell. u

Creating a surface using kriging interpolation with a variable radius

1. Click the Spatial Analyst dropdown arrow, point to Interpolate to Raster, and click Kriging. 2. Click the Input points dropdown arrow and click the point dataset you wish to use. 3. Click the Z value field dropdown arrow and click the field you wish to use. 4. Click the Kriging method you wish to use. 5. Click the Semivariogram model dropdown arrow and click the model you wish to use. 6. Click the Search radius type dropdown arrow and click Variable. 7. Optionally, change the default number of points. 8. Optionally, specify a maximum distance. 9. Optionally, change the default Output cell size. 10. Optionally, check Create Prediction of standard error. 11. Specify a name for the Output raster or leave the default to create a temporary dataset in your working directory. 12. Click OK.

2 3 4 5 6 7 8 9 W Q E
147

PERFORMING SPATIAL ANALYSIS

With a Fixed radius, the radius of the circle used to find input points is the same for each interpolated cell. The default radius is five times the cell size of the output raster. By specifying a minimum number of points, you can ensure that within the Fixed radius, at least a minimum number of input points will be used in the calculation of each interpolated cell.
Tip Deciding on the radius or the number of points Use the Measure tool on the Tools toolbar to measure distance between points to get an idea of the radius and number of points to use. Tip Changing the lag size, major range, partial sill, and nugget Click Advanced Parameters on the Kriging dialog box to specify these parameters if they are known; otherwise, Spatial Analyst will estimate them for you.

Creating a surface using kriging interpolation with a fixed radius

1. Click the Spatial Analyst dropdown arrow, point to Interpolate to Raster, and click Kriging. 2. Click the Input points dropdown arrow and click the point dataset you wish to use. 3. Click the Z value field dropdown arrow and click the field you wish to use. 4. Click the Kriging method you wish to use. 5. Click the Semivariogram model dropdown arrow and click the model you wish to use. 6. Click the Search radius type dropdown arrow and click Fixed. 7. Optionally, change the default distance for the search radius setting. 8. Optionally, change the minimum number of points. 9. Optionally, change the default Output cell size. 10. Optionally, check Create Prediction of standard error. 11. Specify a name for the Output raster or leave the default to create a temporary dataset in your working directory. 12. Click OK.

2 3 4 5 6 7 8 9 W Q E
USING ARCGIS SPATIAL ANALYST

148