Introduction to Bayesian Methods Using the Thomas‐Fiering Model

Investment decisions in water resources planning depend on estimation of unknown parameters. The Bayesian approach to estimation involves use of the estimator that minimizes the expected risk, where risk is the expected loss. Bayes' estimators may be used in a wide variety of situations for which probabilistic models have been specified. In an example, a Bayes estimator of the serial correlation coefficient in streamflow is shown to be a weighted average of the sample serial correlation coefficient and the mean of all population serial correlation coefficients.