An ensemble-based assessment of bias adjustment performance, changes in hydrometeorological predictors and compound extreme events in EAS-CORDEX

Abstract

The effectiveness of adaptive measures tackling the effects of climate change is dependent on robust climate projections. This becomes even more important in the face of intensifying extreme events. One example of these events is flooding, which embodies a major threat to highly vulnerable coastal urban areas. This includes eastern Asia, where multiple coastal megacities are located, e.g. Shanghai and Shenzhen. While the ability of general circulation models (GCMs) and regional climate models (RCMs) to project atmospheric changes associated with these events has improved, systematic errors (biases) remain. This study therefore assess capabilities of improving the quality of regional climate projections for eastern Asia. This is performed by evaluating an ensemble consisting of bias adjustment methods, GCM-RCM model runs and future emission scenarios based on representative concentration pathways (RCP) obtained from EAS-CORDEX. We show that bias adjustment significantly improves the quality of model output and best results are obtained by applying quantile delta mapping. Based on these results we evaluate potential future changes in crucial hydrometeorological predictors, univariate extreme events and compound extreme events, focusing on high wind speeds and extreme precipitation. Key findings include an increase in daily maximum temperature of 1.5 to nearly 4 °C, depending on the scenario, as well as increased levels of precipitation under RCP 8.5. Furthermore, a distinct intensification of extreme events including high temperatures and heavy precipitation is detected and this increase exceeds the increase of the overall mean of these predictors. The annual number of compound events including heavy precipitation and extreme wind speeds shows a significant increase of up to 50% for RCP 8.5 in the South China Sea as well as the adjacent coastal areas.