Short Term Rain and Flow Prediction for Small Tropical Watersheds

In this project an attempt has been made to develop a model to predict the rainfall and flow for small watersheds. The model is applied to two typical tank systems of Kothamangalam tank of Pudukkottai district and Aralikottai tank of Pasumpon Muthuramalinga Thevar district of Tamilnadu, India. Rainfall and flow prediction,adapting the Venkataraman single time series approach and Wiener multiple time series approach were conducted for these systems. To simplify the process, a package was developed in computer using C language for initial statistical analysis, Venkatraman autoregressive scheme and Wiener multiple time series. The prediction results of Aliyabad (control) are compared with that of the tank systems. The result indicated that the raw prediction of daily values are closer to actual values than Trend identified predictions. The sister seasonal time series were more amenable for prediction than whole parent time series. The Venkataraman single time approach was more suited for rainfall prediction. Wiener approach proved better for daily prediction of flow based on rainfall.