Global urban activity changes from COVID-19 physical distancing restrictions.

During the COVID-19 pandemic changes in human activity became widespread through official policies and organically in response to the virus's transmission, which in turn, impacted the environment and the economy. The pandemic has been described as a natural experiment that tested how social and economic disruptions impacted different components of the global Earth System. To move this beyond hypotheses, locally-resolved, globally-available measures of how, where, and when human activity changed are critically needed. Here we use satellite-derived nighttime lights to quantify and map daily changes in human activity that are atypical for each urban area globally for two years after the onset of the pandemic using machine learning anomaly detectors. Metrics characterizing changes in lights from pre-COVID baseline in human settlements and quality assurance measures are reported. This dataset, TRacking Anomalous COVID-19 induced changEs in NTL (TRACE-NTL), is the first to resolve COVID-19 disruptions for all metropolitan regions globally, daily. It is suitable to support a variety of post-pandemic studies that assess how changes in human activity impact environmental systems.