中文摘要：

采用相关分析方法，研究了6～8月长江下游强降水频数与前期冬季12～2月全球50hPa纬向风的关系；并研制了基于KLM滤波的夏季长江下游强降水频数气候预测模型。结果表明，存在3个显著的相关区：北极附近，北太平洋中纬度和南半球高纬度地区。用这些前期12～2月全球平流层关键区的纬向风建立KLM预测模型进行预测试验，发现它对近10年6～8月长江下游地区强降水频数的变化具有较好的预测能力，明显优于多变量的线性回归预测；其中与北极附近和南半球高纬度地区平流层纬向风相关稳定，而与北太平洋中纬度纬向风的相关存在一定程度的不稳定。基于平流层环流变化的夏季长江下游强降水频数KLM模型可以为短期气候预测业务预报提供重要参考。

英文摘要：

Using correlation analysis, the relationships of the number of the heavy precipitation over lower reaches of the Yangtze river valley （LYRV） in summer to the global zonal wind at 50 hPa during previous winter are explored. And the climate prediction model with the KLM filter technique is developed. The results show that there are three regions with significant correlation between the number of the heavy precipitation over LYRV and the global zonal wind at 50 hPa during previous winter, which occur near the NoAh Pole, over the middle latitude of North Pacific, in the high latitude of the Southern Hemisphere. Based on the stratospheric winds of these key regions, the prediction model the number of the heavy precipitation over LYRV in summer can be better predicted for last 10 years, which the prediction for KLM filter technique is significantly better than multiple linear regression prediction. In these connections, correlations to the stratospheric zonal winds of the North Pole and he high latitude of the Southern Hemisphere are stable, and to that over the middle latitude of NoAh Pacific, to some extent, is instable. So, the prediction from KLM filter technique can offfer important to operational prediction of the short term climate prediction for heavy precipitation over LYRV.