Urbanization is projected to increase local surface temperature by 2100

Future projection of global land surface temperature often emphasizes climate change while neglecting urbanization. Yet, urbanization-induced warming strongly influences heatwave-related health risks and energy demands. Here, we developed a 1-km resolution global land surface temperature dataset for 2020–2100 at five-year intervals, combing climate change-induced global warming and urbanization-driven local warming, which were estimated using multi-model ensemble projections, and a dynamic regression model linking impervious surface area and local temperature, respectively. Our dataset aligns closely with satellite observations, showing high spatial and temporal consistency. By 2100, urbanization contributes an average local warming of 0.1 °C, with approximately 10–16% of urban areas experience extreme warming exceeding 1 °C. Urban areas remain warmer than the global mean, whereas their warming rates are 0.5–8% lower than the global average under all scenarios. The derived dataset enables improved assessments of urban heat risks assessments and supports climate-resilient urban planning.