摘要
基于我国科学家自主开发的AREM区域中尺度模式,考虑模式初值和侧边界条件的不确定性,采用BGM方案,建立有11个成员、37km分辨率、覆盖全国区域的短期集合预报系统,并将其于2009年汛期投入准业务化试验,进行了连续2个月的预报结果检验。结果表明:由于能提供更多预报信息,集合预报系统优于单一确定性预报,集合平均预报对评分的改进主要表现在暴雨以下量级,概率预报高于集合平均和控制预报,因而更具参考价值;利用多种集合产品对降水过程进行综合分析,能提高决策预报的科学性。
Based on a regional model AREM (Advanced Regional Eta Model) and its initial conditions and lateral boundary conditions, a Short-Range Ensemble Forecast system which has 11 members in 37km horizontal resolution over China is structured by using BGM scheme. The system was run daily in a quasi operation and evaluated for two months during flood season in 2009. The results show that ensemble forecasts have an advantage over the single deterministic forecasts with additional forecast information. As for the Ts score the ensemble average forecast can improve forecast accuracy of precipitation with a lower rainfall grade. It shows that probability forecast seems to be better than ensemble average forecast and control forecast. Furthermore, the decision-making forecast will be more scientific by analyzing various ensemble products comprehensively.
出处
《暴雨灾害》
2010年第1期30-37,共8页
Torrential Rain and Disasters
基金
科技部公益性行业(气象)科研专项(GYHY200706012
GYHY200906010)资助
武汉暴雨研究所科研业务项目(0603
0903)共同资助
关键词
短期集合预报
AREM模式
降水预报
结果检验
Short-range ensemble forecast
AREM model
Precipitation forecast
Result verification