Decadal Prediction Using a Recent Series of MIROC Global Climate Models

• DOI: 10.2151/jmsj.2012-a22
• Publication Date: 2012-06-14T22:35:54Z
• Authors: Takashi MOCHIZUKI; Yoshimitsu CHIKAMOTO; Masahide KIMOTO; Masayoshi ISHII; Hiroaki TATEBE; Yoshiki KOMURO; Takashi T. SAKAMOTO; Masahiro WATANABE; Masato MORI
• Keywords: Hindcast; Predictability; Initialization; Forcing (mathematics)

In line with the experimental design for near-term climate prediction toward 5th Assessment Report of Intergovernmental Panel on Climate Change (IPCC-AR5) and Coupled Model Intercomparison Project Phase 5 (CMIP5), we perform ensembles initialized decadal hindcast experiments using two recent versions Interdisciplinary Research On (MIROC): MIROC4h (T213L56 AGCM 1/6-1/4 deg. 48 level OGCM) MIROC5 (T85L40 0.56-1.4 50 OGCM). We analyze sets 10-yearlong 9-ensemble hindcasts (3 members by 6 MIROC5) initialization every five years after 1961 explore predictability changes.The most predictable variation timescales is global warming signal due to favorable response models external forcing. The results these validate our ability enhance primarily through initialization, particularly Pacific Decadal Oscillation (PDO) a few Atlantic Multidecadal (AMO) almost decade. has large impacts upper ocean temperature over mid-and high latitudes North Atlantic, where PDO AMO signals are observed be strongest. contribution process assessment studies in IPCC-AR5 CMIP5, further analysis data (and data) worthwhile. note that have predictive skill inferior at this stage, fully significant discussions may not possible small number limited computational resources.