Stochastická simulace denních srážkových úhrnů v povodí řeky Bečvy

Abstract

The aim of this thesis was to test and compare different interpolation and simulation methods, especially conditional stochastic simulation to estimate spatial daily precipitation. Testing was carried out for the Bečva catchment distrikt. In the first section, the precipitation is described, its origin, method of measurement using a network of raingauge stations and also their processing by the Czech Hydrometeorological Institute. The next section provides information on the Becva catchment. The hypothesis is processed by analyzing the current situation. An important part of the thesis is exploratory analysis, assessment of normality and the correlations with the precipitation. The spatial dependence of daily precipitation was described by the chosen model of variogram. To test the selected method the SGeMS modeling software was used, designed for geostatistical analysis of the spatial tasks. The results of estimations and simulations were evaluated using a root mean squared error, bias and conditional or kriging variance. In conclusion, the results obtained were evaluated and the possibility of using SGeMS software for these purposes was assessed.