The inverse modelling technique seeks to improve the existing estimates of natural recharge in hard rocks by coupling multiple hydrogeophysical parameters that jointly affect natural hydrogeological processes. This approach involves coupling of an initial set of multiple hydrogeophysical (soil resistivity, bedrockdepth and rainfall) parameters in the form of exponents assigned to each parameter and a multiplication coefficient to obtain natural recharge. These model parameters (i.e., exponents and coefficients) are then quantified using linear least squares inversion against the known recharge values. To reduce the effect of geomorphic heterogeneity, viz., hills on natural recharge, laterally constrained inversion has been employed to integrate data sets (e.g., recharge measured at various points and logical expectation over exposed hills in an area) and constrained interpolation is then carried out along the grid lines for increaseddata density. Finally, Kriging interpolation over dense data obtained through data integration and constrained interpolation is used to significantly minimise the risks of overshooting the observations. Thus, the present approach provides a realistic spatial distribution of natural recharge values in a highly heterogeneous hard rock terrain.