Automatic evaluation of stochastically generated rainfall data

Abstract

The Cooperative Research Centre for Catchment Hydrology (CRCCH) has embarked on a research project under the Climate Variability Program to build a tool kit that will include stochastic models to generate rainfall data. Stochastic models for generating rainfall data are usually evaluated subjectively by comparing a number of parameters estimated from the historical and the generated data. Models generally use a set of parameters to generate data and these are referred to as the basic parameters. However, there are other parameters or characteristics of the data that also need to be preserved if the generated data is considered to acceptably represent the historical data. A good generated dat..