TR-25 : Surface fitting of groundwater table by means of least square approach

Abstract

With the advent of high speed computers, the model studies for the analysis of complex groundwater flow problems, have become a common practice. For carrying out such model studies, the special and temporal variation of water table levels and aquifer parameters, in a systematic pattern of space coordinate is imperative as an input to the model which are amenable to numerical differentiation. But the water table elevations and aquifer parameters are measured only at few points in space, where an observation well is located. Thus the available historical record of water-table and aquifer parameters are point values from which the spatial distribution in a systematic pattern is required to be assessed. Using the spatial variation of these values the interpolation of water-table levels at nodal points and average aquifer parameters values for a finite difference. grid is to be assessed for conducting model studies.

This report deals with such interpolation of water table levels at nodal points, using least square technique to fit a trend surface or regression on surface coordinates approximating the water table levels. For such surface fitting of appropriate degree, computer program has been developed and a simple test problem was solved.