空气质量与我们生活息息相关，近年来，大气环境质量问题日益严重，对人类的生活与健康产生重要的影响。气溶胶污染是大气环境污染的一项重要来源，是影响大气环境质量的主要因素之一。其中，气溶胶光学厚度（Aerosol Optical Depth，AOD）作为气溶胶的重要参数，对研究气溶胶污染具有重要意义。目前常用的各种气溶胶光学厚度获取手段虽然各具优势，但是普遍存在着数据资料缺失严重或时空分辨率低等缺陷，因此为气溶胶污染的监测带来众多不便。随着全球导航卫星系统（Global Navigation Satellite System，GNSS）的全面发展，其应用空间得到了极大的扩展，应用潜能也在各方面得到充分发挥。大气环境质量与人类的生产生活及健康息息相关，然而大气环境质量不但与污染源有关，气象条件也对大气环境和人体健康有重要的影响。大气气溶胶污染作为大气污染的重要形式之一，同样会受到气象因素的影响。目前，已有学者研究发现气溶胶光学厚度与气象参数之间确实存在着密切关系。因此，为深入研究气溶胶污染对空气质量的影响，提升气溶胶光学厚度数据资料的完整性具有重要研究意义。故本文结合GNSS系统高精度、高时空分辨率、不受天气影响及数据资料完整等多种优势以及多系统GNSS数据可以提升单系统测量数据的有效性、稳定性及精确度的特点进行探究GNSS水汽产品在AOD预测研究中的应用，并初步探究GNSS水汽资料及AOD在实际应用中的可行性。本文主要研究内容如下：

（1）利用小时分辨率ERA5气象数据结合中国大陆构造环境监测网络（Crustal Movement Observation Network of China，CMONOC）的小时分辨率天顶总延迟（Zenith Total Delay，ZTD）数据计算了小时分辨率的大气可降水量（Precipitable Water Vapor，PWV）。通过无线电探空数据（Radio Sounding，RS）、ERA5 PWV等多种数据源对本文反演所得GNSS PWV精度进行验证。结果表明，GNSS PWV的实际精度为2.25 mm且通过对其日变化分析发现与气候事件具有一定响应，进一步证实本文计算所得GNSS PWV具有较高精度。

（2）分析了小时分辨率GNSS PWV、气压及气温与AOD之间的相关性，发现在京津冀地区GNSS PWV、温度与AOD之间均存在正相关关系，相关性分别为0.60与0.38；而气压与AOD之间为负相关关系，相关性为-0.34。上述结果表明，AOD与气象因子之间存在着相关关系。基于这一结论，提出一种基于GNSS PWV及气象参数的AOD自适应预测模型，通过与传统的多元线性回归模型和非自适应的预测模型进行对比验证所提出的AOD自适应预测模型的精度。结果表明，本文所提出的AOD自适应预测模型精度均优于多元回归模型和非自适应预测模型，其中AOD自适应预测模型外符合精度的均方根误差（Root Mean Squared Error，RMSE）和平均绝对误差（Mean Absolute Error，MAE）分别为0.17和0.14。

（3）以新冠肺炎事件为背景，从周末效应、不同人口等级城市、新冠肺炎事件下长时序异常及空间异常等多方面分别进行分析GNSS PWV及AOD的变化特征。结果表明，GNSS PWV及AOD能够很好的反映出周末效应及不同人口等级城市下的变化趋势且与实际现象相一致。此外，GNSS PWV及AOD的长时序与空间异常变化特征与新冠肺炎这一突发事件发生的时间点具有一致性，进一步证明GNSS PWV及AOD的变化趋势与人类活动的剧烈程度有密切关系且在新冠肺炎这一突发公共事件背景下GNSS PWV与AOD具有实际应用的能力。

Air quality is closely related to our life. In recent years, the problem of atmospheric environmental quality has become increasingly serious, which has an important impact on human life and health. Aerosol pollution is an important source of atmospheric environmental pollution and one of the principal factors affecting atmospheric environmental quality. Among them, aerosol optical depth (AOD) is one of the critical parameters of aerosols, which is of great significance to the study of aerosol pollution. At present, although various commonly used aerosol optical depth acquisition methods have their own advantages, they generally have the defects of a serious lack of data or low temporal and spatial resolution, which brings much inconvenience to the monitoring of aerosol pollution. With the comprehensive development of Global Navigation Satellite System (GNSS), its application space has been greatly expanded, and its application potential has been brought into full play in all aspects. Atmospheric environment quality is related to human life and health. However, atmospheric environmental quality is not only related to pollution sources, but also meteorological conditions have an important impact on the atmospheric environment and human health. As one of the principal forms of air pollution, aerosol pollution will also be affected by meteorological factors. At present, scholars have found that there is a close relationship between aerosol optical depth and meteorological parameters. Therefore, in order to deeply study the impact of aerosol pollution on air quality and improve the integrity of aerosol optical depth data, it is of huge significance. Therefore, this paper explores the application of GNSS water vapor products in AOD prediction research by combining the advantages of GNSS system, such as high precision, high spatiotemporal resolution, no weather influence and complete data, as well as the characteristics that multi-system GNSS data can improve the effectiveness, stability and accuracy of single system measurement data, and preliminarily explores the feasibility of GNSS water vapor data and AOD in practical application. The main research contents of this paper are as follows:

(1) Based on the hourly resolution meteorological data of ERA5 and the hourly resolution zenith total delay (ZTD) data of Crustal Movement Observation Network of China (CMONOC), the hourly resolution Precipitable Water Vapor (PWV) is calculated. The accuracy of GNSS PWV retrieved in this paper is verified by radio sounding (RS), ERA5 PWV and other data sources. The results show that the actual accuracy of GNSS PWV is 2.25 mm, and through the analysis of its daily variation. It is found that it has a certain response to climate events, which further prove that the GNSS PWV calculated in this paper has high accuracy.

(2) The correlation between hourly resolution GNSS PWV, pressure, temperature and AOD is analyzed. It is found that there is a positive correlation between GNSS PWV, temperature and AOD in Beijing Tianjin Hebei region, and the correlation is 0.60 and 0.38 respectively; there is a negative correlation between pressure and AOD, and the correlation is -0.34. The above results shows that there is a correlation between AOD and meteorological factors. Based on this conclusion, an AOD adaptive prediction model based on GNSS PWV and meteorological parameters is proposed. Accuracy of the proposed AOD adaptive prediction model is verified by comparing with the traditional multiple linear regression model and non adaptive prediction model. The results show that the accuracy of the AOD adaptive prediction model proposed in this paper is better than that of multiple regression model and non adaptive prediction model. The external coincidence accuracy RMSE and MAE of AOD adaptive prediction model are 0.17 and 0.14 respectively.

(3) Based on novel coronavirus pneumonia events, the changes of GNSS PWV and AOD were analyzed from the weekend effect, different population level city, long time sequence anomaly and spatial anomaly. The results show that GNSS PWV and AOD can reflect the weekend effect and the change trend of cities with different population levels, and are consistent with the actual phenomenon. In addition, the characteristics of long sequence and spatial anomaly of GNSS PWV and AOD are consistent with the time of occurrence of new crown pneumonia. It further proves that the trend of GNSS PWV and AOD is closely related to the intensity of human activities. And the ability of GNSS PWV and AOD to be applied in the background of the novel coronavirus pneumonia.