A fuzzy logic model for estimation of groundwater recharge

Water budget models are commonly accepted methods for estimating the groundwater recharge. Some of the inputs to these models such as soil-moisture deficit, actual evapotranspiration, direct runoff etc, are difficult to measure. The space dynamic nature of these parameters makes their quantification even more difficult. The error in estimation increases significantly during extreme events. In addition, it has been found that water budget model is not sensitive to recharge phenomena that occur during intense rainstorms of short duration. This phenomenon is typical during summer in Indian conditions. In this study, a fuzzy rule based model to estimate the groundwater recharge has been presented for fluvially deposited sedimentary area lying south of the foot hills of Himalayas, in and around Roorkee (29/spl deg/ 53" N, 78/spl deg/ 53' 52"E, and 269.120 m), Uttaranchal, India. The model inputs are air temperature, rain fall and stream flow. The mean difference in this estimate and the water budget estimate is about 25%. It is expected that the prediction efficiency can be improved by training the model using a larger historical data set.