A Nonparametric Copula Based Bias Correction Method for Statistical Downscaling

Yi Li, Northeastern University

Global Climate Models (GCMs) currently provide coarse resolution outputs which preclude their application to accurately assess the effects of climate change on finer regional scale events that are important to inform stakeholders in making policy, management, or infrastructure planning decisions. Statistical downscaling are methods that use statistical models to infer the regional-scale or local-scale climate information from coarsely resolved climate models. One popular approach for statistical downscaling is the bias correction and spatial disaggregation (BCSD) method. BCSD utilizes quantile mapping to perform bias correction between the coarse resolution climate models to the fine resolution projection. In this paper, we analyze BCSD from a copula point of view and show that it is a restricted form of the copula function. Instead, we propose a nonparametric copula based BCSD method (NCBCSD) and empirically show that this more flexible method provides improved climate projection performance in terms of mean-squared-error compared to the traditional BCSD method.

Link to Presentation: CI2016 Li