Spatial interpolation of groundwater levels by kriging

The command areas of canal irrigation projects in Rajasthan face the problem of rising groundwater level and manifestation of waterlogging at many places. To delineate the extent of waterlogged area, contour mapping of groundwater levels is required, which in turn require the interpolation of groundwater levels, measured in field at random points, at the grid nodes. In this paper, the applicability and usefulness of kriging to interpolate and map the spatial behaviour of groundwater levels has been demonstrated through a case study. Kriging is applied to interpolate the groundwater levels (above mean sea level) measured at 67 points in pre-monsoon season and at 61 points in post-monsoon season of 1991 in parts of the command area of Indira Gandhi Nahar Pariyajana (IGNP). With the use of measured elevations of the water table (above mean sea level), experimental semivariograms were constructed for both the seasons. Spherical, exponential and gaussian theoretical semivariogram models were fitted to the experimental semivariograms. The finally selected models were used to estimate the groundwater levels and corresponding standard deviation at the nodes of a square grid of 5km x 5km and to develop corresponding contour maps. The effect of using a common semivariogram, parameters estimated by taking the arithmetic mean of the semivariogram parameters for June and September data sets, on the cross validation results is also investigated.