Uncertainty evaluation related with the fitting of probability distributions to rainfall experimental data

Abstract

This study addresses the uncertainty evaluation of the fit of the probability functions (PDF) Exponential, Gamma and Gumbel to precipitation data, describing the method and the related mathematical models and using known procedures to obtain the uncertainty of the PDF parameters. Experimental data obtained from a climatological station is used. The method of Moments is applied to estimate the parameters of each fit PDF and their uncertainties are evaluated using a Monte Carlo Method. The obtained results show the differences obtained regarding the impact of the measurement uncertainty of the input quantities (precipitation) and similarities related with the shape of the parameters PDF.

Keywords

• Uncertainty
• Precipitation
• Probability distribution
• Rainfall data