Optimal parameter identification of nonlinear, time - variant hydrologic system models

This study focuses on the important need in river basin investigations for development of mathematical models that describe hydrologic system response to rainfall. In particular time variant, nonlinear stimulated models for runoff and a predictive model for evapotranspiration are presented. One of the major objectives of this study is in the development of a hydrological model for water delivery particularly suited for a developing nation, though general in application. Emphases are placed on using what is available in the best possible way. The proposed model is evaluated with respect to total computation time required and ease of understanding and utilization. Solution procedures are evaluated with respect to speed of convergence to optimal solutions of computed runoff and parameters; and general accuracy and efficiency. Two models are developed; first and second order nonlinear differential equations. From the results obtained, the first order nonlinear differential equation proves to be a better monthly model for river basin response stimulation. This conclusion is based on computed runoff using the models, and the predicted seasonal and annual evapotranspiration losses.