Seasonal characteristrics of air-sea coupled process and their implication on predictability

Atmosphere-ocean coupled process is investigated in the vein of the crucial source of low predictability in tier-two forecast system. Both in monthly and pentad data, it is clearly shown that the wrong reproducibility of seasonally varying air-sea interaction induces the lower predictability, in particular over the western North Pacific region. To understand the mechanism of seasonally varying atmosphere-ocean interaction, various prediction systems including AGCM, slab ocean mixed-layer model, and CGCM are compared. For the focus on the thermodynamic process, it is investigated that the local air-sea interaction is mainly modulated by shortwave flux over the summer hemisphere where shows negative relationship with heavy rainfall and cloud cover. While, the role of latent heat flux on the local air-sea interaction is more important in winter hemisphere. Except the regions where ocean dynamics strongly operates such as tropical central and Eastern Pacific, this thermodynamic process is valid for dominant mechanism to determine the direction of air-sea interaction.