The Influence Of Tropical Intraseasonal Oscillation On Pre-flood Season Precipitation In South China And Its Application In Extended-range Forecast

Based on precipitation data on stations, Real-Time Multivariate Madden-Julian oscillation (MJO) index, boreal summer intraseasonal oscillation, outgoing longwave radiation (OLR) data, CMAP precipitation data, NCEP/NCAR reanalysis data, the influence of the Madden-Julian oscillation and boreal summer intraseasonal oscillation on pre-flood season precipitation in South China and its possible mechanism is investigated by composite analysis. A statistical extended-range forecast model of pre-flood season precipitation in South China based on tropical intraseasonal oscillation was established by singular value decomposition (SVD). The result suggests that:(1) There are significant differences in the precipitation in South China between MJO phases, and these differences vary with the lagged length of time. MJO phase2-3and phase6-7correspond to wet phase and dry phase, respectively. The MJO response in pre-flood season precipitation in South China is on a time scale of about1-2pentads. The precipitation anomaly is a result of a Rossby wave associated with the enhanced (suppressed) convection of the MJO. When MJO enhanced (suppressed) convection locates at Indian Ocean, the Rossby wave excited by diabatic heating reaches South China and increases (decreases) water vapor transportation. Such a progress contributes to the enhancement (reduction) of the precipitation in South China.(2) There are also significant differences in the precipitation in South China between BSISO phases. BSISO phase2,3,4,5,6and phase7,8,1,2correspond to wet phase and dry phase, respectively. When MJO enhanced (suppressed) convection propagates northward and reaches South China and Bay of Bengal, the convection causes upward (downward) motion anomaly and increased (decreased) water vapor transportation. Such a progress also contributes to the enhancement (reduction) of the precipitation in South China.(3) Based on the relationship between tropical ISO and the precipitation in South China, together with the periodicity of ISO, a statistical extended-range forecast model of pre-flood season precipitation in South China was established by singular value decomposition. This model forecast2-6pentads precipitation in South China by past1-5pentads RMM and BSISO index. The model reflects the the relationship between tropical ISO and the precipitation in South China. The precipitation index reforecasted exceeds95%confidence level.