2014 Volume 92 Issue 3 Pages 207-225
Evaluations of the summer/winter Asian monsoon through the late 20th century (1981-2000) were conducted on the basis of model simulations using 20 Coupled Model Intercomparison Project Phase 3 (CMIP3) and 24 Phase 5 (CMIP5) multi-model datasets, and comparisons of the results with many types of observational data. Skill metrics have been calculated in terms of reproducibility of seasonal mean structures. The projected thermal structure of the mid to upper troposphere, which is an important driving force of the Asian monsoon, was also evaluated. Overall, the skills of the CMIP5 multi-model ensemble (MME) mean results have been improved, as compared with those of the CMIP3 MME.
Considering these evaluations, we examined projected future (2081-2100) changes in the summer/winter Asian monsoon, including those of the tropical Hadley-Walker circulation, for mid-range emission scenarios (SRES-A1B for CMIP3 and RCP4.5 for CMIP5). The CMIP3 MME shows projected increases in precipitation and attenuation of circulation over broad regions of Asia. This so-called “wind-precipitation paradox” is a characteristic property of the Asian monsoon under a CO2-rich atmosphere. The CMIP5 MME, on the other hand, shows a projected acceleration of climatological low-level monsoon westerlies, particularly in subtropical regions (10°-20°N), which therefore requires a partial revision of the wind-precipitation paradox. In terms of meridional temperature gradients (MTGs), the CMIP5 MME datasets project marked mid to upper tropospheric warming over the western Indian Ocean, as compared with other regions of the Indian and western Pacific oceans. At higher latitudes, the projected warming rate is relatively small to the northwest of the Tibetan Plateau, and projected MTGs are reduced in this region. In the summer Asian monsoon, the different circulation change between CMIP3 and CMIP5 MME despite the common MTG weakening is a notable feature.
