Derivations of simple site-specific recharge-precipitation relationships: A case study from the Cuddalore Basin, India

Abstract

Population growth and higher living standards have resulted in an ever-increasing demand for water. Quantifying the current rate of groundwater recharge is a prerequisite for efficient and sustainable groundwater resource management in dry areas. The normal practice of assuming natural recharge value as certain percentage of rainfall may be suitable for yearly estimation, but this method miserably fails if it is used for short-term estimation. Equations were developed for natural recharge estimation and applied to a case study for Cuddalore Basin, Tamilnadu, India. Natural recharge was estimated using water level fluctuation method, water balance method, nonlinear regression method, and artificial neural network (ANN) model. According to the water level fluctuation method, on an average, 18.74% of the rain got recharged to the ground. The average annual recharge to the groundwater reserve was 18.98, 16.81, and 18.48% of rainfall in the case of the water balance model, nonlinear regression model, and ANN model. The study shows that the framework adopted for the nonlinear regression and ANN models can be locally used to predict natural recharge based on precipitation estimates.