[A-CG52] Large Ensemble Modeling Approaches as Tools for Climate and Impacts Research

[ACG52-02] Event attribution using large ensemble model simulations by MIROC5-AGCM, MRI-AGCM, and NHRCM

★Invited Papers

*今田 由紀子¹、川瀬 宏明¹、塩竈 秀夫³、森 正人⁴、高橋 千陽²、荒井 美紀⁵、渡部 雅浩²、高薮 出¹ (1.気象庁気象研究所、2.東京大学大気海洋研究所、3.国立環境研究所、4.東京大学先端科学技術研究センター、5.海洋研究開発機構)

キーワード：イベント・アトリビューション、Large ensemble simulation、極端気象現象

There are many evidences of connections between long-term trends of extreme events and the human influence. However, specific extreme events can not totally be blamed on the past human activity, because extreme events can be happened only due to the natural variability. In this study, we examine whether human influence has largely affected the probability of extreme events by comparing large-ensemble climate model simulations under the "factual" conditions and the "counterfactual" ones in which influences of anthropogenic climate drivers were omitted. Such an attempt is called "Event Attribution" (EA).
We conducted large ensemble simulations for the factual and counterfactual experiments using two atmospheric general circulation models: MIROC5-AGCM and MRI-AGCM3.2. The MIROC5 dataset is superior for consideration of the range of uncertainty, whereas the MRI-AGCM3.2 dataset, so called d4PDF (the Database for Policy Decision making for Future climate change), focuses on high-resolution simulations with ~60 km horizontal resolution along with downscaled products produced by a regional climate model NHRCM with ~20km horizontal resolution which covers the East Asian region.
Here, I will introduce our recent works using the large ensemble simulations focusing on how much anthropogenic global warming contributed to increased occurrence of recent extreme events around the globe when compared to the inter-annual and decadal variability.