内蒙古地区沙尘和霾自动观测数据的可靠性和修正方法研究*

2014年,中国气象局将地面人工观测业务调整为自动仪器观测的方式,这一改变明显提高了工作效率,但对观测结果,特别是视程障碍类观测产生了一定的影响。为了完善内蒙古地区霾和沙尘天气判别标准,获得准确可信的气象资料,本文选取了2001-2019年内蒙古12个站点逐三小时的气象观测资料,从时空分布特征和判别方法研究了内蒙古地区沙尘和霾自动气象观测数据的可靠性和判别标准的适用性。结合PM2.5和PM10两个空气质量指数,通过修正错误记录,将沙尘中误判的霾和无天气、霾中误判的沙尘、雾和无天气等分离出来。对比结果来看,修正前,霾和沙尘分别占总频次的95%和5%；修正后,霾、沙尘、雾和无天气分别占总频次的55%、17%、1%、27%。对比呼和浩特单站一次沙尘和一次霾天气过程中气象要素和空气质量指数的连续变化可以看出,相对湿度与PM2.5的相关性均达到了0.5以上。沙尘期,能见度与PM10负相关,PM2.5和PM10相关性较差；而霾期,能见度与PM2.5和PM10的相关性较差,PM2.5和PM10相关性达到了0.8以上,即不同相对湿度环境下影响能见度的因素不同。

Reliability and correction method of automatic observation data of dust and haze in Inner Mongolia

In 2014, China Meteorological Bureau adjusted the ground artificial observation business to the automatic instrument observation mode, which significantly improved the work efficiency. But it had a certain impact on the observation results, especially the observation of the apparent range obstacles. In order to improve the criteria of haze and dust weather in Inner Mongolia and obtain accurate and reliable meteorological data, by selecting the meteorological observation data of 12 stations in Inner Mongolia from 2001 to 2019, the reliability of the automatic meteorological observation data of dust and haze and the applicability of the criteria from the spatial-temporal distribution characteristics and discrimination methods are discussed. Combined with PM2.5 and PM10 two air quality indexes, the mistakenly judged haze and no weather in dust and mistakenly judged dust, fog and no weather in haze are separated by correcting the wrong records. Before the correction, haze and dust accounted for 95% and 5% of the total frequency respectively; after the correction, haze, dust, fog and no weather accounted for 55%, 17%, 1% and 27% of the total frequency respectively. By comparing the continuous changes of meteorological elements and air quality index in the process of dust and haze in Hohhot station, it can be seen that the correlation between relative humidity and PM2.5 is more than 0.5.In the process of sand dust, visibility is negatively correlated with PM10, and PM2.5 and PM10 are poorly correlated; in the process of haze period, visibility is poorly correlated with PM2.5 and PM10, and the correlation between PM2.5 and PM10 is more than 0.8, which means that the factors affecting visibility are different under different relative humidity environment.