Climate Simulation over Japan in Summer with an MRI regional climate model: Present Day and 2xCO2 Experiments

Regional climate change simulations in summer ( June, July, August) associated with global warming have been performed with a regional climate model nested in the results of a transient CO2 experiment by a Meteorological Research Institute coupled atmosphere-ocean general circulation model (MRI-CGCM) at a resolution of 5 degrees in longitude and 4 degrees in latitude. The MRI regional climate model is composed of two models; a 120km-resolution East Asian model and a 40km-resolution Japanese Islands model. For long time simulation, the model adopted a spectral boundary coupling in addition to a conventional boundary relaxation nesting.

Two kinds of transient CO2 experiment output from the CGCM are used in these simulations. One is a control run with a fixed atmospheric CO2 concentration at 345 ppmv (1xCO2). The other is a run with a gradually-increased CO2 concentration at a compound annual rate of 1%. For each of the month in summer, ten month-long simulations using the regional model were carried out for these two cases from 70th when atmospheric CO2 concentration doubles (2xCO2).

Under the 1xCO2 condition, the regional climate model reproduces more realistic climate than that of the CGCM; in particular; regional features of monthly-averaged surface air temperature are represented. The nested regional climate model represents a Baiu front which is a typical climate feature in East Asia during the summer monsoon season. Moreover, the simulation of the detailed spatial pattern of the precipitation in August due to thermal low is clearly improved. However, maximum precipitation is overestimated for these areas.

Differences(warming signal) of the surface air temperature between two cases(2xCO2-1xCO2) are ranges from 2 to 2.5 degrees on Japan, which are larger than biases between 1xCO2 run and observed climate. As to the precipitation, warming signal for each month shows that the amount of precipitation tend to decrease on the Japan and increase on the south Pacific ocean area near Japan. However, it is difficult to obtain significant quantitative results because biases are larger than the warming signal.

These results show that regional climate prediction has a certain degree of feasibility though in this model there are still some physical processes such as surface hydrological cycle including vegetation effects to be improved.