Water management regulates vegetation recovery in photovoltaic farms across China

PV (photovoltaic) farms are rapidly expanding across China, affecting its land use patterns and terrestrial ecosystems. While previous studies have shown that the effects of PV farms on vegetation are diverse, regulated by land cover, climate, and technology type, the role of human interventions, especially water management, in influencing vegetation dynamics remain insufficiently examined, leaving a critical knowledge gap. To bridge this gap, we assessed the vegetation dynamics of China’s PV farms using the normalized difference vegetation index, estimated human-induced evapotranspiration change (ETh) as a measure of water management intensity, and used machine learning models to assess its impact on vegetation dynamics. Our results showed that water management plays a critical role in driving vegetation recovery, with optimal ETh forming a climate-dependent ‘ridge pattern’ that increased from 92 mm/year in humid regions to 200 mm/year in arid zones. Optimized water management can increase the improvement in vegetation greenness to 2.3 times that under observed scenario, while reducing total water consumption by 21.5%. Importantly, our findings provide evidence-based guidance for region-specific water management policies aiming to promote ecological restoration in PV farms while also supporting sustainable land use practices.