Robust Bayes Models of Fish Response to Lake Acidification

摘要

A large historical data set from the Adirondack region of the United States was created to study the relationship between water chemistry variables associated with acid precipitation and the presence/absence of selected fish species. From these data, logistic function models for presence/absence were fitted for brook trout and lake trout, using maximum likelihood and robust Bayes methods. The robust Bayes procedure used herein is a compromise estimator between Bayes and empirical Bayes, with the final expression based on classical risk function criteria. A cross-validation study will be used to aid in selection of a preferred model for policy analysis.