The COVID-19 pandemic provided an unprecedented natural experiment to assess vegetation responses to reduced anthropogenic pressures. We developed a multi-scale framework synthesizing policy stringency, human mobility, and remote sensing data to quantify pandemic impacts on global vegetation. Specifically, we constructed a spatially explicit Human Modification Stringency Index to capture relative shifts in human activity from 2017 to 2023, while characterizing vegetation dynamics using Dynamic World land cover and MODIS NDVI. Our results reveal that global vegetation area, which had declined by ∼2.3 million km 2 before the pandemic, rebounded with a net gain of ∼2.9 million km 2 during the pandemic. This reversal was corroborated by growing-season NDVI, which transitioned from browning (−0.0021) to greening (+0.0056). Using a continuous Difference-in-Differences model across ∼28.3 million pixels, we causally attributed this greening to reduced human disturbance in 96.6% of significant grid cells. The effects peaked in 2022 before diminishing as restrictions eased. Notably, moderately human-modified landscapes showed the most widespread recovery, whereas heavily modified areas exhibited the greatest magnitude of change, suggesting substantial latent vegetation recovery potential. These findings highlight that targeted reductions in anthropogenic pressures can drive rapid vegetation recovery, offering critical insights for ecological restoration and land management. • Global vegetation increased by ∼2.9 million km 2 during the COVID-19 pandemic. • Novel Human Modification Stringency Index quantified pressure reductions. • Continuous DID model decoupled anthropogenic effects from climatic drivers. • Greening was most prevalent in semi-modified but greatest in heavily modified areas.
He et al. (Sun,) studied this question.