Uncovering patterns of converging human-induced pressure on global lands

Monitoring how human activities and other environmental pressures jointly affect different parts of the Earth is critical for guiding conservation, sustainable development, and climate adaptation efforts. Yet, high-resolution, temporally consistent tools to track these combined pressures globally remain limited. Using the newly developed machine learning Human Footprint Index version 2 (ml-HFIv2.0), we conduct an analysis of the intersection between human pressure on land and climate change. The ml-HFIv2.0 was developed using a convolutional neural network trained on Landsat satellite data, building upon a previous version. This new version features a finer spatial resolution of 300 meters and the ability to maintain temporal continuity using Landsat imagery as input. These methodological advancements result in a fully self-sufficient workflow, enabling retrospective and up-to-date assessments of human pressure on global land. Moreover, ml-HFIv2.0 effectively captures rapidly evolving land-use changes, such as illegal mining activities, that were previously overlooked by conventional datasets. Our analysis shows that natural areas, which cover most of the world, exhibit disproportionately higher warming, with 58.5% of these regions warming faster than the global median. In contrast, land areas experiencing some degree of human pressure have typically experienced slower warming, with around 65% of these areas warming at a rate below the global median. Approximately 2.7% of global land area has experienced both high human pressure and rates of warming above the global median. Our findings underscore the importance of analyzing human and climatic pressures simultaneously, highlighting how different land-use categories coincide with global warming patterns. The ability of ml-HFIv2.0 to provide high resolution, up-to-date monitoring offers critical advantages for understanding and managing anthropogenic impacts and their interactions with climate change. These insights are essential for informing conservation, urban planning, and climate adaptation strategies.