Chen, H. and Sun, J. 2013. Projected change in East Asian summer monsoon precipitation under RCP scenario. Meteorology and Atmospheric Physics 121: 55-77.
The authors write that "over East Asia, the climate is dominated by the summer monsoon in the rainy season," where the associated rain belt is called the Meiyu in China, the Baiu in Japan, and the Changma in Korea. And noting that "the advance and retreat of the Meiyu belt result in a large variability of precipitation over East Asia and generally lead to many natural disasters, e.g., floods and droughts," they go on to state that "projection of future changes in precipitation and its frequency and intensity is a critical issue for long-term planning for national and regional adaptation and mitigation," implying the need for superlative climate modeling in this regard.
What was done
The ultimate objective of Chen and Sun was to investigate future changes in precipitation frequency and intensity over the East Asian region; but before they could accomplish that task, it was necessary for them to evaluate the degree to which the coupled ocean-atmosphere general circulation models (CGCMs) from CMIP5 reproduce the present-day climate, which they thus proceeded to do.
What was learned
First of all, the two researchers found that "most of the models in CMIP5 underestimated the amount of precipitation over China, Korea and Japan," and that some of them even "lacked the Meiyu belt over this region." In addition, they say that "because of the underestimation by most of the models, the MMEs (multi-model ensembles) of all of the models also showed a negative bias." Secondly, a number of the models additionally showed "almost no skill in simulating the precipitation frequency." And they add that "similar to the single models, the frequency was also overestimated by the MMEs in the high- and low-latitude regions of East Asia." Last of all, they report that "most of the models under-estimated the precipitation intensity over eastern China, Korea, Japan and their adjacent oceans."
What it means
Chen and Sun conclude that "similar to the results based on CMIP3 models, CMIP5 models also show no skill in simulating precipitation variation with time," and, therefore, they say that "more research is needed in the future to further improve the reliability of climate projections." Or, looking at it from a different perspective, maybe we should just give up on the fruitless and very expensive enterprise.