An Empirical Bayes Framework for Assessing Changes in the Hydrological Cycle

Abstract:

Greenhouse gases in the atmosphere have been increasing since the industrial revolution, leading to the warming of the Earth through an increase in downwelling infrared radiation. Warming of the atmosphere increases its water holding capacity and could intensify the hydrological cycle. Several methods have been developed for evaluating changes in climatic variables. On the other hand, numerous indices have been developed for monitoring changes in climatic variables. Most change detection methods, indices, and trend studies focus on changes in one variable at the time. However, hydrologic variables are dependent, and a change in one variable can alter extreme and non-extreme values of other variables. In this study, a new approach for modeling multivariate extreme values through a conditional distribution framework using the empirical Bayes approach is proposed. This study highlights the value of empirical Bayes conditional extreme value analysis as a tool for simulating and assessing conditional extremes (e.g., changes in the distribution of precipitation conditioned on extreme temperature). The model has been applied to several locations across the world. This presentation will summarize the findings on changes in the hydrological cycle over the United States and Australia.