Abstract:

Fog and haze are disaster weathers which endanger human health and affect social and economic development. Accurate and detailed monitoring data can play an important role in the prevention and control of fog and haze. The accuracy of China Meteorological Administration Land Data Assimilation System (CLDAS) visibility and relative humidity fusion products in identifying fog, light fog and haze is analyzed by using the observation data of national stations in Tianjin and its surrounding areas from December 1, 2017 to November 30, 2020, Himawari-8 L1 full-disk data and L3 aerosol optical depth product. The results show that compared with the station observation data, the average detection rates of CLDAS products in identifying light fog, fog and haze are 90.4%, 84.2% and 78.8%, respectively. The detection rates of light fog in different months are 81.1%-96.4%. In the months with more fog and haze, the detection rates are about 80.0%. The cases analysis shows that the fog, light fog and haze identified by CLDAS products are basically consistent with the results of Himawari-8 satellite and observations. The failure of CLDAS products to correctly identify fog, light fog and haze mainly shows that fog is misjudged as light fog (3.8%-21.4% at different stations) and haze is missed (8.6%-25.0% at different stations). When the horizontal visibility of the station is between 0 and 0.75 km, the error of CLDAS visibility mainly causes fog to be mistakenly identified as light fog. When the horizontal visibility of the station is between 0.75 and 7.5 km，the error of CLDAS visibility mainly leads to haze being missed. When the station visibility is between 7.5 and 15 km, the error of CLDAS visibility mainly leads to light fog and haze being reported empty. When the relative humidity of the station is greater than 40% and less than or equal to 60%, the error of CLDAS relative humidity mainly leads to haze being misjudged as light fog. In general, the accuracy of CLDAS products in identifying fog, light fog and haze in Tianjin area is good, which can provide reference for fine monitoring of fog, light fog and haze, and improve the status quo of scarce visibility observation stations and insufficient space coverage in fog and haze monitoring.

Key words: CLDAS products, fog and haze, visibility, relative humidity, accuracy identification

摘要：

雾和霾是危害人类健康和影响社会经济发展的灾害天气，精细化的实况资料能够在雾和霾的防治中发挥重要作用。利用2017年12月1日至2020年11月30日天津及其周边地区国家气象观测站资料、Himawari-8卫星L1级全圆盘观测数据和L3级气溶胶光学厚度产品，分析了中国气象局陆面数据同化系统（CMA Land Data Assimilation System，CLDAS）能见度和相对湿度融合实况分析产品判识天津地区雾、轻雾和霾的准确性。结果表明：与台站资料相比，CLDAS产品对轻雾、雾和霾的平均检出率分别为90.4%、84.2%和78.8%；CLDAS产品对轻雾的逐月检出率为81.1%~96.4%，雾和霾出现较多的月份，其检出率均在80.0%左右。个例分析表明CLDAS产品判识的雾、轻雾和霾与台站观测结果以及Himawari-8卫星反演检测结果基本一致。CLDAS产品未正确判识雾、轻雾和霾的情况主要表现为雾误判为轻雾（各站为3.8%~21.4%）和霾漏判（各站为8.6%~25.0%）。当台站水平能见度在区间[0，0.75 km）时，CLDAS能见度的误差主要导致雾误判为轻雾；在区间[0.75，7.5 km）时，CLDAS能见度的误差主要导致霾漏判；在区间[7.5，15 km）时，CLDAS能见度的误差主要导致轻雾和霾空报。当台站相对湿度大于40%且小于等于60%时，CLDAS相对湿度的误差主要导致霾误判为轻雾。总体而言，CLDAS产品对天津地区雾、轻雾和霾的判识准确性较好，能够为雾、轻雾和霾的精细化监测提供参考，改善雾和霾监测中能见度观测站点稀少、空间覆盖不足的现状。

关键词: CLDAS产品, 雾和霾, 能见度, 相对湿度, 准确性判识