| Abstract: | The aim of this paper is to present a rainfall generator which takes both spatial and temporal characteristics into account. The statistical model behind the generator is based on long-time experience with statistical analyses of empirical rainfall data and represents, therefore, a genuine theoretical point for simulating series of precipitation values. The sites of the recording gauges and the actual time (day, month, year etc.) enter a statistical model as systematic components for the prediction of expected rainfall in a two-way analysis of variance design. Beyond these systematic components, spatial correlations among the observations are still present in nature, a feature which is modelled also by the rainfall generator. |
