引用本文:蔡沅辰,丁峰,朱志锋,张良瑜,李源慧,徐振麒.南京市空气质量预报效果评估及误差分析[J].环境监控与预警,2023,(2):28-32
CAI Yuan-chen,DING Feng,ZHU Zhi-feng,ZHANG Liang-yu,LI Yuan-hui, XU Zhen-Qi.Evaluation and Error Analysis on the Effect of Air Quality Forecast in Nanjing[J].Environmental Monitoring and Forewarning,2023,(2):28-32
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  【EndNote】   【RefMan】   【BibTex】
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 2607次   下载 678 本文二维码信息
码上扫一扫!
分享到: 微信 更多
南京市空气质量预报效果评估及误差分析
蔡沅辰,丁峰,朱志锋,张良瑜,李源慧,徐振麒1,2
1.江苏省南京环境监测中心,江苏 南京 210013;2.南京信息工程大学,环境科学与工程学院,大气环境与装备技术协同创新中心,江苏省大气环境监测与污染控制高技术研究重点实验室,江苏 南京 210044
摘要:
基于2020年南京市空气质量实况数据及预报数据,评估了当年南京市空气质量预报效果,分析了预报偏差特征及其成因。结果表明,4个季节中,秋季的空气质量指数(AQI)预报准确率评分和综合评分最高,夏季的首要污染物准确率评分最高;4个季节均出现正预报偏差,其中夏、冬季偏差大于春、秋季;首要污染物误报率与季节相关,二氧化氮(NO2)和可吸入颗粒物(PM10)的误报率较高的原因是NO2和PM10作为首要污染物主要出现在春、秋季,而这2个季节4种主要污染物的空气质量分指数(IAQI)值非常接近,增加了预报员经验修正的难度。典型预报偏差个例分析结果表明,模式预报对于污染物质量浓度量级的预报偏差以及预报员对气象条件和前体物质量浓度变化关注不足,是导致最终预报出现低估的主要原因。
关键词:  空气质量预报  空气质量指数  效果评估  误差分析  区域气象-大气化学在线耦合模式
DOI:
分类号:X823
基金项目:国家自然科学基金项目(42177211);江苏省环境监测科研基金项目(2112);南京环保科技项目(202111,202016)
Evaluation and Error Analysis on the Effect of Air Quality Forecast in Nanjing
CAI Yuan-chen,DING Feng,ZHU Zhi-feng,ZHANG Liang-yu,LI Yuan-hui, XU Zhen-Qi
1.Jiangsu Nanjing Environmental Monitoring Center, Nanjing, Jiangsu 210013, China
Abstract:
Using the observed and forecast data of air quality in Nanjing in 2020, this study evaluated the air quality forecast effect in Nanjing in 2020, and analyzed the characteristics and causes of forecast bias. The results showed that among the four seasons, the AQI forecast accuracy score and comprehensive score are the highest in autumn, while the accuracy score of primary pollutants is the highest in summer. Positive prediction deviations occur in all four seasons, and the deviations in summer and winter are greater than those in spring and autumn. The false alarm rate of pollutants is related to the season. The reason for the higher false alarm rate of NO2 and PM10 is that, as the primary pollutants, NO2and PM10 mainly appear in spring and autumn. Because the IAQI values of the four pollutants in these two seasons are very close, it increases the difficulty of the forecaster subjective correction. The results of the forecast deviation analysis of a typical case showed that prediction bias of model for pollutant concentration and the insufficient attention of the forecaster to the changes of meteorological conditions and precursor concentrations cause the underestimation of the final forecast.
Key words:  Air quality forecast  AQI  Effect evaluation  Error analysis  WRF_chem