IDENTIFICATION OF RUNOFF MODELS FOR NONPOINT POLLUTANT LOAD IN URBAN AREAS

Abstract

In this paper, we present a parameter identification procedure for the pollutant load runoff models in urban areas. The procedure consists of two steps: the parameter estimation based on the optimization method and the observation noise separation of the input-output data with the neural network model. This procedure is effective for most of the pollutant load runoff models in urban areas and expected to give a more reliable and stable model parameter estimation.