Future projections and uncertainty assessment of precipitation extremes in the Korean peninsula from the CMIP5 ensemble

Abstract

Projections of changes in extreme climate are sometimes predicted by multimodel ensemble methods that combine forecasts from individual simulation models using weighted averaging. One method to assign weight to each model is the Bayesian model averaging (BMA) in which posterior probability is used. For the cases of extreme climate, the generalized extreme value distribution (GEVD) is typically used. We applied the approach of GEV-embedded BMA to a series of 35 years of the annual maximum daily precipitation data (both historical data and data gathered from simulation experiments for future periods) over the Korean peninsula as simulated by the models in the Coupled Model Intercomparison Project Phase 5 (CMIP5). Simulation data under two Representative Concentration Pathway (RCP) scenarios, namely RCP4.5 and RCP8.5, were used. Observed data and 17 CMIP5 models for 12 gird cells in Korea have been examined to predict future changes in precipitation extremes. A simple regional frequency analysis of pooling observations from three stations in each cell was employed to reduce the estimation variance and local fluctuations. A bias correction technique using the regression-type transfer function was applied to these simulation data. Return levels spanning over 20 and 50 years, as well as the return periods relative to the reference years , were estimated for two future periods, namely Period 1 (2021-2050) and Period 2 (2066-2095). From these analyses, relative increase observed in the spatially averaged 20-year (50-year) return level was approximately 23% (16%) and 45% (36%) in the RCP4.5 and RCP8.5 experiments, respectively, by the end of the 21st century. We concluded that extreme rainfalls will likely occur two times and four times more frequently in the RCP4.5 and RCP8.5 scenarios, respectively, as compared to in the reference years by the end of the 21st century. K E Y W O R D Sclimatic change, global climate model, heavy rainfall, multimodel simulation, Taylor diagram, weighted averaging

Changes in temperature and rainfall extremes across East Asia in the CMIP5 ensemble

“…Seo et al (2015) obtained that the values decrease to 5 and 4 years relative to historical data (11 and 8 years relative to observations) for RCP4.5 and RCP8.5, respectively. Lee et al (2020) reported these values be 10 and 5 years relative to the observations in 1971-2005, which is similar to the result of this study. Table 2 summarizes changes (%) in the CMIP5 median and quartiles (25% and 75%) relative to 1986-2005 in zonally averaged R 20 across the globe, EA, and Korea in 2081-2100.…”

“…The ∆R 20 over EA and Korea are higher than the global figures reported in K13. The result in Korea is comparable with those by Lee et al (2020), in which they reported 23% and 45% ∆R 20 increases for RCP4.5 and RCP8.5 respectively, relative to the observations in 1971-2005.…”

“…Korea shows similar sign of the projected changes with enhanced trend. The 20-year waiting times for Korea are comparable with the results by Seo et al (2015) and by Lee et al (2020). Seo et al (2015) obtained that the values decrease to 5 and 4 years relative to historical data (11 and 8 years relative to observations) for RCP4.5 and RCP8.5, respectively.…”

“…To solve this problem, one can apply the bias correction technique which constructs a mapping that matches the observations and the historical data well. Some studies, including Piani et al (2010), Sangelantoni et al (2019), Shin et al (2019a), and Lee et al (2020), applied the statistical bias correction methods to the output of models, in which the obtained mapping for the past was applied to the simulation data for future periods. Such application may be valid under the assumption that the relationship between observations and historical simulation is kept very similar in the future.…”

A Weighting Scheme in A Multi-Model Ensemble for Bias-Corrected Climate Simulation

“…In our experience, when non-bias-corrected (non-BC) historical data are used in BMA, only a few models exhibit extreme weights and most others have very low weights [25] (see Figure 3). Specifically, the posterior distribution may excessively depend on a few "outlier models" close to the observation, when all other models fail to capture observations of the historical period-a common situation for precipitation metrics [26].…”