基于时空统计降尺度的淮河流域夏季分月降水概率预测

针对淮河流域水资源短缺、洪涝、干旱并存的问题,基于国家气候中心第二代季节气候模式的集合回报数据集(1991—2014年),建立时空相结合的统计降尺度模型,提前1—3个月预测该流域夏季分月降水,应用ROC(relative operating characteristics)评分评估比较了不同集合预测方案的预测技巧。交叉检验结果表明,样本数取18、20、22、28时,集合预测方案对3、4、5月三个起报时次预测的夏季各月降水技巧预测均高于模式预测技巧。2015—2017年的独立样本检验进一步表明该统计降尺度模型能够明显降低3月、5月起报的6月和8月的降水预测偏差。认为可尝试将该降尺度方法应用于淮河流域夏季降水预测及进一步的流域水文预测。

Probability Prediction of Monthly Precipitation over Huaihe River Basin in China in Summer Based on Spatio Temporal Statistical Downscaling Method

Skillful precipitation prediction is useful in managing water resources and preventing droughts and floods for the Huaihe River Basin. A hybrid spatio temporal statistical downscaling model based on the ensemble reforecast dataset (1991-2014) of the seasonal climate model (CSM1.1m) was developed to predict the summer monthly precipitation over the basin ahead of 1-3 months, i.e., starting from March, April and May. Cross validation tests indicate that compared with CSM1.1m, the established statistical downscaling model is more skillful for the precipitation forecasted in March, April and May for four ensemble schemes. Independent sample tests present that the statistical downscaling model can reduce prediction error, especially for precipitation in June and August forecasted in March and May. These results suggest that it is possible to use the statistical downscaling method for precipitation probability forecasting and further hydrological forecasting in summer over Huaihe River Basin.