[A-CG32] Climate Variability and Predictability on Subseasonal to Centennial Timescales

[ACG32-05] Subseasonal to seasonal scale prediction with JMA/MRI-CPS3 and its future challenges

★Invited Papers

*Shoji Hirahara¹, Yutaro Kubo^1,2, Takuma Yoshida^1,2, Takuya Komori^1,2, Toshinari Takakura^1,2, Hiroyuki Sugimoto^1,2,3, Yukimasa Adachi^2,1, Ichiro Ishikawa^1,2, Yosuke Fujii^1,2, Yuhei Takaya^1,2 (1.Meteorological Research Institute, 2.Numerical Prediction Development Center, 3.Japan Meteorological Agency)

Keywords:seasonal forecast system, predictability

Numerical seasonal forecasting systems of the Japan Meteorological Agency (JMA) have evolved over more than two decades to meet growing societal demands of climate information. This presentation introduces a new operational seasonal forecast system, JMA/ Meteorological Research Institute (MRI) Coupled Prediction System (CPS) version 3 (JMA/MRI–CPS3) and its future prospects. JMA/MRI-CPS3 consists of an atmosphere, land, ocean, and sea ice forecast models and initialization systems required for these models. The ocean and sea ice models are initialized using a new version of a global ocean data assimilation system (MOVE-G3) with a four-dimensional variational scheme for temperature, salinity and sea surface height and a three-dimensional variational scheme for sea ice concentration. Compared to the previous system, improved initialization methods, higher model resolution as well as refined physical processes, have resulted in improved forecast skill for subseasonal to seasonal scale variability such as the Madden–Julian Oscillation, winter blocking highs, and El Niño–Southern Oscillation. However, there remain some notable shortcomings, such as excessive development of the Indian Ocean Dipole, which limits the forecasting capability. The possible causes and side effects of these biases are discussed.