Global high-resolution mapping of photovoltaic power plants from 2019 to 2025 using unsupervised index-based multi-source data fusion method

• An efficient framework enables rapid and low-cost global PV plant mapping. • The adaptive index-based data fusion method ensures more accurate PV extraction. • The first 10 m-resolution global PV mapping product from 2019 to 2025. • The mapping product delivers the most up-to-date global PV plant distribution. • Spatio-temporal PV diffusion and land-use changes are systematically revealed. Accurate mapping of photovoltaic (PV) power plants is critical for monitoring the development of solar energy generation and supporting PV operational management, policy-making, and environmental analysis. However, existing geospatial datasets of PV power plants are either limited in spatial coverage, insufficient in resolution and accuracy, or outdated and lack timely updates, thus failing to meet these demands. To address this, we develop an efficient framework for global PV mapping that integrates the proposed adaptive normalized difference photovoltaic index (ANDPI) with a multi-source data fusion algorithm to extract accurate contours of PV power plants from Sentinel-2 imagery. We generate the global mapping product of PV power plants at 10 m resolution from 2019 to 2025. This mapping achieves an overall accuracy of 91.16 %, outperforming existing PV mapping products, while maintaining the processing speed of 140 km 2 per second. Our analysis reveals that the global PV area expanded from 8432 km 2 in 2019 to 16289 km 2 in 2025, with China contributing 61 % of the growth. Besides, PV power plants are primarily deployed on rangeland (44.0 %) and cropland (40.1 %). Cropland held a slight dominance before 2021, after which rangeland accounted for a larger proportion. To the best of our knowledge, this geospatial dataset represents the most comprehensive global PV mapping product, offering the global spatial coverage, the highest resolution, and the most up-to-date information. It provides valuable insights into the spatial distribution and expansion of PV power plants, contributing to global strategies for carbon neutrality and sustainable development. The code and data are available at https://github.com/SUNNYTREE-LAB/Fusion_PV_Mapping.