The seasonal predictability of East Asian Summer monsoon

Abstract:

The Fifth Coupled Model Intercomparison Project (CMIP5) set up the near-term climate prediction experiments which provide seasonal-to-decadal information for earth system. The seasonal prediction skill for East Asian summer monsoon (EASM) is a long-standing climate issue, but has not reached a consensus due to its complicated internal variation and external forcing. To address the seasonal predictability of EASM, we focused on the mean squared error skill score (MSESS), the correlation, the conditional bias and the continuous ranked probability skill score of five prediction systems (CanCM4,GFDL-CM2.1, HadCM3, MIROC5 and MPI-ESM-LR) for the June-July-August temperature and precipitation during 1960-2009. The results show (1) that temperature has strong potential prediction skill on seasonal-to-inter-annual scale; (2) that initialisation simulations improve the prediction skill compared to non-initialisation simulations; (3) that strong monsoon years exhibit better performance in seasonal prediction than weak monsoon years.