Investigation of effective parameters in estimation of groundwater table using kriging with an external drift method

Estimation of groundwater table has always been very important, especially nowadays that groundwater resources are considered as one of main water supply resources. The kriging with an external drift (KED) method is a useful tool that can be used to improve the estimation of groundwater table. This is due to the fact that, in addition to the available piezometric data as the primary data, topography data as the secondary data are used for external drift in the estimation of groundwater table. In this paper, the sensitivity of the KED method has been examined with respect to the correlation coefficient between primary and secondary data and the portion of nugget effect in the structure of primary data. The aim of this study is to attain some criteria for use of the method and to test its capability in different conditions. The sensitivity analysis of the method has been carried out based on the cross validation results. The investigated parameters in this sensitivity analysis consist of the average estimation error, the average estimation error variance, slope of fitted regression line over estimated and actual data, and also the determination coefficient obtained from this regression. Finally, the sensitivity of the KED method for each of the above-mentioned parameters has been carried out and the results of this sensitivity analysis have been presented and interpreted in this paper.