Geostatistics and Hydrology. Part 2. Estimation Techniques. Communications of theSanitary Engineering and Water Management Division

摘要

In hydrology a linear estimator is often used when the value at an unmeasuredpoint is estimated. This type of estimator is based on the assumption that the data measured near the estimated location are likely to be in more correspondence with the value to be estimated. Measured data close to the estimated location receive a higher weight than measured data at a larger distance. A whole, more or less arbitrary, range of methods is given among which is Kriging. In this method the weights are selected on the basis of the spatial structure of the phenomenon, using a covariance function. Once the covariogram has been developed it is used to estimate distance weights for interpolation. The weights are selected so that estimates are unbiased and with minimum variance.