著者
YUKIMOTO Seiji KAWAI Hideaki KOSHIRO Tsuyoshi OSHIMA Naga YOSHIDA Kohei URAKAWA Shogo TSUJINO Hiroyuki DEUSHI Makoto TANAKA Taichu HOSAKA Masahiro YABU Shokichi YOSHIMURA Hiromasa SHINDO Eiki MIZUTA Ryo OBATA Atsushi ADACHI Yukimasa ISHII Masayoshi
出版者
Meteorological Society of Japan
雑誌
気象集誌. 第2輯 (ISSN:00261165)
巻号頁・発行日
pp.2019-051, (Released:2019-06-18)
被引用文献数
441

The new Meteorological Research Institute Earth System Model version 2.0 (MRI-ESM2.0) has been developed based on previous models, MRI-CGCM3 and MRI-ESM1, which participated in the fifth phase of the Coupled Model Intercomparison Project (CMIP5). These models underwent numerous improvements meant for highly accurate climate reproducibility. This paper describes model formulation updates and evaluates basic performance of its physical components. The new model has nominal horizontal resolutions of 100 km for atmosphere and ocean components, similar to the previous models. The atmospheric vertical resolution is 80 layers which is enhanced from 48 layers of its predecessor. Accumulation of various improvements concerning clouds, such as a new stratocumulus cloud scheme, led to remarkable reduction in errors in shortwave, longwave, and net radiation at the top of the atmosphere. The resulting errors are sufficiently small compared with those in the CMIP5 models. The improved radiation distribution brings the accurate meridional heat transport required for the ocean and contributes to a reduced surface air temperature (SAT) bias. MRI-ESM2.0 displays realistic reproduction of both mean climate and interannual variability. For instance, the stratospheric quasi-biennial oscillation can now be realistically expressed through the enhanced vertical resolution and introduction of non-orographic gravity wave drag parameterization. For the historical experiment, MRI-ESM2.0 reasonably reproduces global SAT change for recent decades; however, cooling in the 1950s through the 1960s and warming afterward are overestimated compared with observations. MRI-ESM2.0 has been improved in many aspects over the previous models, MRI-CGCM3/MRI-ESM1, and is expected to demonstrate superior performance in many experiments planned for CMIP6.