10-30天延伸期预报是介于常规天气预报和短期气候预测之间的预报"缝隙"，其可预报性指预报对象具有10-30天潜在可预报时效，且能提前10-30天找到与预报对象有一定联系并超过一定显著性的可预报信号。不同时空尺度的梅雨降水过程可预报时效有多长？提前10-30天是否存在来自热带和中高纬的周期性、持续性大气可预报信号？延伸期可预报信号影响不同时空尺度降水过程的机理如何？这些问题的回答有助于加深对延伸期可预报性的认识。因此，本项目拟采用非线性信号提取方法EEMD和非线性误差增长理论，结合区域气候模式RegCM4，系统分析江淮梅雨季节不同时空尺度降水过程在10-30天延伸期的可预报时效和前期大气信号，揭示前期信号影响降水过程可预报性的机理，构建不同条件下的延伸期信号影响降降水过程的物理模型，为延伸期预报提供科学支撑。

10-30dy extended range forecast (ERF) is a forecast "seam" between tradional weather forecast and short-term climate prediction. The predictability of ERF means that predictants have 10-30dy potential predictability and significant predictable signals relating to predictant can be found at the same time. How long are the preidictable periods of Meiyu rainfall processes with diferent temoral or spatial distribution? Is there any atmospheric signals with periodicity, persistence in 10-30dy advance? What are mechanisms of influences of atmospheric signals on differencet rainfall processes? Answers to these questions may be helpful to intensify the knowledge about predictability of ERF. By multiple utilization of nonlinear signal detecting method EEMD,nonlinear error growth theory and regional climate model RegCM4, the project will systematically investigate predictable period and atmospheric signal of rainfall processes with different temporal and spatial distribution, and reveal the mechanisms of atmospheric signal's influence on Meiyu rainfall processes, thus construct physical model for influences of atmospheric signals on rainfall processes under different large scale conditions to provide scientific basis for ERF.