• 关键词：PM_2.5  CMAQ模式  北京  2018年  模式评估

• 基金项目：国家重点研发计划（No.2017YFC0209805，2017YFC0212100）；中央高校基本科研业务费专项资金资助项目；北京高精尖学科"陆地表层学"资助

• 李冬青
• 北京师范大学全球变化与地球系统科学研究院, 北京 100875

• 吴其重
• 北京师范大学全球变化与地球系统科学研究院, 北京 100875

• 徐旗
• 北京师范大学全球变化与地球系统科学研究院, 北京 100875

• 王晓彦
• 中国环境监测总站, 国家环境保护环境监测质量控制重点实验室, 北京 100012

• 高超
• 中国人民解放军 63921部队, 北京 100094

• 肖晗
• 北京师范大学全球变化与地球系统科学研究院, 北京 100875

• 王蓉蓉
• 北京师范大学全球变化与地球系统科学研究院, 北京 100875

• 摘要：自北京奥运会以来，CMAQ模式作为北京多模式预报系统的一个成员在北京空气质量预报中得到广泛应用.为了更好地发展模式系统预报性能，本文针对北京开展完整年份PM_2.5模拟预报效果评估，结合我国环境空气质量标准，引入IAQI预报准确率、等级预报准确率以及预报综合评分法等多项指标，研究评价不同代际不同模式分辨率CMAQ模式系统预报效果差异.研究结果表明，①新研发的CMAQ模式系统预报效果整体优于原有业务预报模式系统，综合考虑预报级别准确性和预报空气质量分指数精确度的得分评估结果显示，新一代CMAQ模式3 km空间分辨率（BJ03）4 d预报时效内得分为73.1~80.5分，高于9 km空间分辨率（BJ09）和原有CMAQ模式5 km空间分辨率（CN05）和15 km空间分辨率（CN15）的预报结果综合得分.②BJ03区域的预报效果优于BJ09区域，BJ03区域预报的PM_2.5-IAQI准确率达42%，空气质量等级准确率达68%~79%，预报综合评价得分最高80.5分；而9 km分辨率CMAQ模式预报的PM_2.5-IAQI准确率为30%，等级准确率为53%~70%，预报综合评价得分最高为75.3分.③模式PM_2.5-IAQI预报准确率整体随预报时长增加而下降，原有业务系统CMAQ模式的CN05、CN15区域IAQI预报准确率由24 h内的27%下降至7 d预报时长的22%左右；BJ03区域IAQI预报准确率由24 h内的42%下降到4 d预报时长的29%，与之对应的BJ09区域IAQI预报准确率由24 h内的30%下降到4 d预报时长25%，至9 d预报时长，IAQI预报准确率进一步下降到22%，即达到原有业务预报系统7 d IAQI预报准确率.④模式秋冬季的PM_2.5预报结果整体较实况偏高，BJ03区域秋冬季较实况偏高9.2 μg·m^-3，而BJ09、CN05和CN15区域较实况偏高25.6~37.9 μg·m^-3.

• Abstract：CMAQ model, as a member of the Beijing multi-model prediction system, has been widely used in the Beijing air quality prediction since the Olympic Games in 2008. This study evaluated the model capabilities with diverse resolutions of predicting PM_2.5 in Beijing for year 2018 using the multi-index methods based on the national environmental air quality standard, e.g., IAQI prediction accuracy, grade prediction accuracy, and prediction comprehensive scoring method, for better improving the prediction performance of CMAQ in the future. The results show that: ① the performance of the new developed CMAQ model system with higher resolution is better than that of the original operational forecasting model. Based on the combination of the forecast level accuracy and the prediction accuracy of IAQI, 3 km spatial resolution of new CMAQ (BJ03) has a range of 73.1 to 80.5 during four-day forecast period, which is much higher than that of 9 km spatial resolution (BJ09) and original model system with 5 km(CN05) and 15 km(CN15). ② The forecast ability of BJ03 is better than that of BJ09. The prediction accuracy rate of PM_2.5-IAQI, the accuracy rate of corresponding air quality level and the highest score of BJ09 is 30%, 53%~70% and 75.3, respectively, which is lower than that of BJ03 of 42%, 68%~79%, and 80.5, respectively. ③ The prediction accuracy rate of PM_2.5-IAQI declined with the increase of prediction duration. For example, the prediction accuracy rate is 42% for 24 hours prediction in BJ03 and decreases to 29% for the 4-days forecasting; while it is 30% in 24 hours prediction and decreases to 25% for 4 days prediction, even 22% for 9 days case, in BJ09. In addition, the prediction accuracy rate decreases from 27% for 24 hours prediction to about 22% for 7 days prediction in original model system. ④ Model forecasting results are relatively higher than observation in autumn and winter. The PM_2.5 concentration result of BJ03 is 9.2 μg·m^-3 higher than the observation, while BJ09, CN05, and CN15 are 25.6~37.9 μg·m^-3 higher than the observation.

• Key words：PM_2.5  CMAQ model  Beijing  2018  model evaluation

• 
  摘要点击次数： 846 全文下载次数： 895