SEAS5模式对新疆月尺度气温和降水的预测性能评估

检验评估气候模式，有利于发挥这一客观预测方法在新疆气候预测业务中的应用效果，提高新疆气候预测能力。基于欧洲中期天气预报中心第五代季节预测系统SEAS5模式的历史回报数据和新疆98个气象站的气温、降水观测资料，利用距平符号一致率Pc、趋势异常综合评分Ps、距平相关系数ACC、时间相关系数TCC等方法，综合评估了SEAS5模式对新疆1993—2019年月尺度气温和降水的预测性能。结果表明：SEAS5模式对新疆月气温、降水的总体预测性能较好，对降水趋势和量级有一定的预测能力。模式的预测性能存在明显的月际差异和空间差异。模式对新疆4—9月气温和4月、6—8月降水的预测性能较好。气温预测评分中，4月Pc最高（61.3分），7月Ps最高（74.4分），6月ACC最高（0.17）。降水预测评分中，4月Pc最高（58.7分），7月Ps最高（72.1分），4月ACC最高（0.16）。空间分布上，模式对1月、3月、6月和12月南疆的气温预测性能普遍高于北疆，对7月和9月南疆西部的预测性能低于其他地区。模式对南疆的降水预测性能普遍高于北疆。

Monthly Temperature and Precipitation Evaluation of SEAS5 in Xinjiang

Evaluating climate models is conducive to give full play to the application of the model. Based on the data of SEAS5 climate model and data of 98 weather stations over Xinjiang, we evaluate the monthly-scale climate prediction ability of the model by calculating Pc, Ps, ACC and TCC scores. The results show that SEAS5 has a good ability to predict the temperature and precipitation in Xinjiang. There are obvious temporal and spatial differences in the performance of the model. The model has good performance for the temperature prediction from April to September. SEAS5 can predict precipitation well from June to August in Xinjiang. In terms of temperature prediction, the Pc score in April is the highest (61.3), the Ps score in July is the highest (74.4), and the ACC score in June is the highest (0.17). In terms of precipitation prediction, the Pc score in April is the highest (58.7), the Ps score in July is the highest (72.1), and the ACC score in April is the highest (0.16). In terms of spatial distribution, the temperature prediction skill of the model in Southern Xinjiang is generally higher than that in Northern Xinjiang in January, March, June and December. The prediction performance for the west of Southern Xinjiang in July and September is lower than that in other regions. The precipitation prediction skill of SEAS5 model in Southern Xinjiang is generally higher than that in Northern Xinjiang.