Paper Title
Best Input Variable Combinations for the Reservoir Capacity Analysis

Abstract
Water yield is affected by precipitation and is impacted by different worldwide climatic parameters. In this study, an expectation of reservoir yield of 'Upper Bhima River Basin ' from the Maharashtra State of India has endeavored. The worldwide climatic input parameters, to be specific El Nino-Southern Oscillation (ENSO) file and Equatorial Indian Ocean Oscillation (EQUINOO) file are utilised for the forecast. The Genetic Programming (GP)and Artificial Intelligence (AI) tools have been utilized through Software ‘Discipulus’ for this purpose. Upper Bhima basin comprises of numerous dams and harnessing water for irrigation, hydro-power and household employment. Inflow, outflow, water use and losses of all reservoirs are taken into consideration while calculating actual yield Upper Bhima River Basin. Five combinations of input variables for predicting reservoir yield were tested to arrive at the best input variable combination for better predictions. From the analysis, it is discovered that models created utilizing GP build up a sensibly great relationship between atmospheric factors (climate variables) and basin yields. The best combination of ENSO and EQUINOO gave co-relationship coefficient r = 0.9726 between observe driver basin yield and anticipated river basin yield, which seems attractive for such a mind-boggling complex system.