The Role of Prior Probabilities on Parameter Estimation in Hydrological Models

Bayes theorem provides a formal framework for combining prior and sample information parameter estimation in the presence of measurement structural errors. Prior knowledge, however, may not always be available or too vague to incorporate into distribution. In such cases, reference must chosen an objective analysis. Typically, uniform density over possible ranges parameters is as prior. However, validity seldom questioned. this study, information-theoretic approach pursued derive priors, results are compared those obtained by using Examples estimating saturated hydraulic conductivity presented. Priors transformation-invariant typically nonuniform. The choice between influences posterior distribution conductivity, when small. use also demonstrated through PDG-GIUH hydrologic model, issues computational tractability addressed.