10:00 〜 10:15
[ACG38-05] High-resolution large ensemble simulation with an ocean-assimilated climate model
キーワード：高解像度気候モデル、大規模アンサンブル実験
In order to obtain more precise future regional climate projection information including the range of uncertainties, a climate change prediction system has been updated based on the CMIP6 participation model of the Meteorological Research Institute (MRI-ESM2). The system is named TSE-C (Time Sequential Experiments with Coupled model). By assimilating water temperature, salinity and sea ice concentration with a relaxation time of about 5-10 days, the bias from observational climatology is suppressed while explicitly simulating short-term atmosphere-ocean interaction. It enables regional-scale climate projection with including short-term atmosphere-ocean interaction such as ocean mixing by overpassing tropical cyclones, lag correlation between precipitation and sea-surface temperature.
The updated large ensemble simulation is now being conducted with 60km atmosphere resolution, continuously from the mid-20th century to the latter half of the 21st century. Since the effects of ocean decadal variability dominates the range of uncertainty until around 2040, an interannual variability phase emsemble, in which different phase of decadal variability of ocean are forced to ensemble members, is combined with a forcing model ensemble, in which different warming patterns of CMIP6 models are forced to ensemble members. That can represent a large fraction of the variability seen among the CMIP6 models. The results are also dynamically downscaled to our regional climate model and ocean model with 20km and higher resolution.
The updated large ensemble simulation is now being conducted with 60km atmosphere resolution, continuously from the mid-20th century to the latter half of the 21st century. Since the effects of ocean decadal variability dominates the range of uncertainty until around 2040, an interannual variability phase emsemble, in which different phase of decadal variability of ocean are forced to ensemble members, is combined with a forcing model ensemble, in which different warming patterns of CMIP6 models are forced to ensemble members. That can represent a large fraction of the variability seen among the CMIP6 models. The results are also dynamically downscaled to our regional climate model and ocean model with 20km and higher resolution.