动力和统计预测信息融合预测方法及对我国夏季降水预测的检验

A merging method for dynamic model and statistical predictions and evaluation of summer precipitation prediction skill in China

短期气候预测中如何将气候模式和统计方法的预测结果科学、客观的集成起来,一直是非常重要的问题.本文针对动力模式和统计方法预测结果相结合的问题,引入资料同化中信息融合的思想,采用最优内插同化方法,实现了动力模式和统计季节降水预测结果的融合.检验表明,对1982—2015年我国夏季降水百分率的回报,融合预测结果与观测的平均空间相关系数可达0.44,分别较统计预测和CFSv2模式统计降尺度订正的技巧提高了0.1左右,而均方根误差较两者可以降低5%~20%.可见,该方法可以进一步提升对我国夏季降水的预测技巧,具有显著的业务应用价值.

How to make an objective combination for the predictions from climate dynamic model and statistical method is an important issue in the short-term climate prediction area. To resolve this problem, we use the optimum interpolation assimilation method to merge the predictions of summer precipitation percentage anomalies in China from CFSv2 model and statistical method. The verification of prediction skill indicates that during 1982-2015 periods, the spatial correlation coefficient of merging predictions with the observations is as highly as 0. 44. which is higher than the skill of statistical prediction method and the downscaling of CFSv2 model about 0. 1. Moreover, the root mean square error of merging predictions is also reduced by 5%~20% than that of statistical method and CFSv2 model downscaling. Our results indicate that this new prediction method can further improve the prediction skill of summer precipitation in China and has the potential application values.