Uncovering relationships between geospatial features and COVID-19 features is a comprehensive, confounding, cross-disciplinary and challenging topic, as the spread and effects of COVID-19 are related to many aspects of our lives, including socio-economic, cultural, and environmental features. Our research aims to provide an innovative data-driven method to uncover the relationships between the heterogeneous and cross-disciplinary geospatial features with COVID-19 features at the municipality scale in Germany. We exploit these relationships using supervised machine learning, explainable AI and spatial analysis in Germany from March 2020 to October 2021. First, we integrated multi-source data including social data, economic data, cultural data, air pollution data and COVID-19 features data into one spatiotemporally harmonised dataset. Second, we trained three machine learning models (a Support Vector Regressor, a Random Forest, and a Light Gradient Boosting Machine) on the integrated dataset to learn the relationships between the spatial features and the COVID-19 features. Third, we used Shapley Additive exPlanations (SHAP) to rank the relevance of each feature. After that, we illustrated the results by the visualised spatial differences within municipalities. The output delivers key information regarding the Covid hospitalisation rate with the control of NO2 concentration and education level in Germany with transferable methods.@en

The outbreak of the coronavirus disease 19 (Covid-19) has posed a worldwide threat to human beings, economic activities, and society. Enforced lockdowns for limiting the spread of Covid-19 virus also substantially reduce air pollutant emissions from vehicle traffic, industrial plants, etc. The lockdown restrictions have brought beneficial environmental implications, such as improvement of air quality. Previous studies recorded the reduction of air pollutants during the short-term lockdown in some cities and areas in Indian, China, and the U.S. [1-5]. While some studies argue that the improvement of air quality is not due to lockdown, but season influence or temporary change by coincidence [6]. Therefore, there is not enough evidence that the improvement of air quality is mainly due to reduced human activities. It is beneficial to answer this question by investigating and comparing the air pollution changes within countries with multi waves of pandemic timelines and different lockdown measures. Our research chose Germany and the Netherlands to investigate the air pollutant changes during their multiple lockdowns. Both of the two countries have gone through several pandemic waves while their imposed strategies are different, ranging from lockdown light, partial lockdown, full lockdown, to curfew in different stages of the pandemic. Our research investigates changes in air quality during their multiple pandemic waves and compares seasonal and monthly changes with the historical records (pre-pandemic) from ground stations to analyze the anomalies. During the pandemic period, the research will compare the disparities of air quality improvement with the several pandemic waves among mega urban agglomerations within the two countries. For the pre-pandemic period, this research analyzes the anomaly in comparison with the historical records with the air quality index. In particular, we adopt the datasets produced by a space-borne air pollution sensor TROPOMI on the Sentinel-5P satellite, provided in Google Earth Engine data catalog. We process the data and extract information about air pollution, including CO, NO2, SO2, O3, and CH4 for analyzing the air pollutant composition changes during the several pandemic waves. First, the decline values of air pollutant composition will be calculated and analyzed between the pandemic waves to prove the different changes following every wave in main urban areas within the two countries. Second, by aggregating the air pollutant concentrations from the satellite-based air pollution data into monthly, seasonal, and annual data and comparing them with corresponding historical records from ground stations at the same periods of pre-pandemic time, the anomalies will be calculated and analyzed to illustrate the improvement of air quality because of pandemic lockdowns at the country level. The historical record data will be collected from the air quality index based on the ground station measurements. Third, the disparities of air pollutant reduction during the pandemic will be also analyzed between the Netherlands and Germany, considering their different lockdown strategies. The result will provide strong evidence on the air quality improvement due to the reduction of human activities during lockdown periods and highlight the influence of anthropogenic activities on air pollution. The resulting information will provide information to policymakers concerning emission control and sustainable urban development. Keywords: Air quality changes, lockdowns, pre-pandemic, Google Earth Engine Reference: 1.Parida, B.R., et al., Impact of COVID-19 induced lockdown on land surface temperature, aerosol, and urban heat in Europe and North America. Sustainable Cities and Society, 2021. 75: p. 103336. 2.Naqvi, H.R., et al., Improved air quality and associated mortalities in India under COVID-19 lockdown. Environmental Pollution, 2021. 268: p. 115691. 3.Berman, J.D. and K. Ebisu, Changes in U.S. air pollution during the COVID-19 pandemic. Science of The Total Environment, 2020. 739: p. 139864. 4.Sahani, N., S.K. Goswami, and A. Saha, The impact of COVID-19 induced lockdown on the changes of air quality and land surface temperature in Kolkata city, India. Spatial Information Research, 2021. 29(4): p. 519-534. 5.Li, L., et al., Air quality changes during the COVID-19 lockdown over the Yangtze River Delta Region: An insight into the impact of human activity pattern changes on air pollution variation. Science of The Total Environment, 2020. 732: p. 139282. 6.Etchie, T.O., et al., Season, not lockdown, improved air quality during COVID-19 State of Emergency in Nigeria. Science of The Total Environment, 2021. 768: p. 145187.@en