DERF2.0模式对贵州延伸期、月预测质量对比分析

利用2014—2016年第二代月动力延伸预测模式业务系统(DERF2. 0)直接输出及预报员预报的贵州省气温和降水预报数据、贵州省84站观测数据,采用Ps、Cs和Zs 3种评分方法评估了DERF2. 0对贵州月气温、降水和月内强降水过程的预测性能,并与预报员业务值班发布的预报质量进行对比分析,系统性地对DERF2. 0产品在延伸期/月尺度的预测效果进行检验评估。结果表明:DERF2. 0月预测产品总体上对贵州气温的预测性能优于降水,气温偏高预测的可参考性高于其对气温偏低的预测。DERF2. 0预报月气温的Ps评分较预报员略高,月降水较预报员略低,总体来看DERF2. 0月预测效果较预报员综合预报略差且预报性能较不稳定。预报员预测数据的Zs和Cs评分总体高于DERF2. 0预测数据,但DERF2. 0预测数据的Zs和Cs评分较预报员逐年提高,尤其在整个雨季时段较为明显。DERF2. 0对主汛期(6—8月)的预报准确率较整个雨季期(4—10月)略高。

Comtrastive analysis on the extended range and monthly forecast quality by DERF2.0 in Guizhou

Based on temperature and precipitation forecast results of the second-generation monthly Dynamic Extended Range Forecast operational system (DERF2.0) and forecasters and the data of 84 meteorological observation stations in Guizhou from 2014 to 2016, temperature and precipitation in the prediction performance were evaluated and analyzed by using Ps, Cs and Zs evaluation scores. And compared with the forecast quality of forecasters, the prediction effect of DERF2.0 were examined and evaluated systematically. The results show that temperature forecast performance of DERF2.0 is better than that of precipitation in Guizhou. The reference of high temperature forecast is higher than that of low temperature. The Ps score of monthly temperature forecast of DERF2.0 is slightly higher than that of forecasters while monthly precipitation is slightly lower than forecasters. Generally speaking, the monthly forecast effect of DERF2.0 is slightly worse than that of forecasters, and the prediction performance is unstable. Zs and Cs scores of forecasters are higher than DERF2.0, but Zs and Cs scores of DERF2.0 increase year by year, especially during the whole rainy season. The prediction accuracy of DERF2.0 in the flood season (from June to Augest ) was slightly higher than that in the whole rainy season (from May to October).