Lab for Aerosol, Radiation, Remote-sensing, and Observation-based Modeling of Atmosphere (ARROMA)


COVID-19 and Atmosphere: A Study of Their Two-Way Interactions

The COVID-19 global pandemic, while still ongoing in many parts of the world, has already resulted in an unprecedented change of atmospheric composition. The ARROMA team (led by Prof. Jun Wang) and Prof. Joe Gomes are collaborating to not only study this change, but also to investigate how weather and atmospheric composition may affect the transmission and survival of COVID-19 virus. Our methods entail a combination of Graph Machine Learning tool (GML), regional chemistry transport model (WRF-Chem), epidemiological (SIR) model for infectious disease, and statistical analysis, while data includes weather and atmospheric composition data from NASA’s Earth Observation System (EOS) and NASA’s Earth Systems Modeling as well as county/state-level daily COVID-19 cases that are now widely available from public health organizations. Below are a few preliminary results.


covid forecast

Distribution of number of COVID-19 cases in the state of New York State. left: foretasted; right: predicted. The length in each segment of the color bar represents the number of counties in that segment. Courtesy: Meng Zhou and Lorena Castro.


covid forecast
Fig.1. TROPOMI NO2 (left) and MODIS (right) during the epidemic (bottom) in 2020, as compared to the same time in 2019 (top). Courtesy: Zhendong Lu and Lorena Castro.

covid forecast
Fig.2. Distribution of number of COVID-19 cases on 4/14/2020 from (a) observation and (b) machine- learning (GML model) prediction. (c) shows correlation analysis (scatter plot) between observed and predicted daily number of cases for 4/8-4/16 2020. All the data are for New York State. The length in each segment of the color bar for (a) and (b) represents the number of counties in that segment. Courtesy: Meng Zhou.

covid forecast
Fig.3. Modeled results of the accumulated infected population and the reproduction number, 𝑅0, as a function of time for China (left) and the state of New York (right). Courtesy: Xi Chen.