Monitoring economic activity using satellite imagery and tracking COVID-19 epidemics impact on air quality

Abstract

Real-time satellite observations combined with machine learning techniques can provide daily snapshots of economic activity. Greenhouse Gas emissions are one possible proxy that can indicates methane, nitrogen dioxide and sulphur dioxide produced by human activity around cities, oil and gas refineries, power and utility plants, and agricultural land. Daily and seasonal local patterns can then be identified which correlate with human activity in combination with weather. With the recent limitations in human mobility triggered by the COVID-19 outbreak, satellites can provide a novel way to remotely monitor economic slowdowns and restart at global scale.