Techniques for estimating uncertainty in climate change scenarios and impact studies

Abstract

Methodology for quantifying uncertainty in global climate change studies is reviewed. The focus is on recent developments in statistics, such as hierarchical modeling and Markov chain Monte Carlo simulation techniques, that could enable more full-fledged uncertainty analyses to be performed as part of integrated assessments of climate change and its impacts. First an overview of uncertainty analysis, including its sources and how it propagates, is provided. Presently employed techniques in climate change assessments, such as sensitivity, scenario, and Monte Carlo simulation analyses, are then surveyed. Next alternative approaches, based on more formal statistical theory (especially the Bayesian statistical paradigm), are described. Finally, some tentative recommendations on strategies for achieving the goal of more reliably quantifying uncertainty in global climate change are made.