Photovoltaic (PV) power generation is one of the world's most promising options for carbon emission reduction. However, whether the operation period of solar parks can increase greenhouse gas (GHG) emissions in hosting natural ecosystems has not been fully considered. Here, we conducted a field experiment to compensate for the lack of evaluation of the effects of PV array deployment on GHG emissions. Our results show that the PV arrays caused significant differences in air microclimate, soil properties, and vegetation characteristics. Simultaneously, PV arrays had more significant effects on CO2 and N2O emissions but a minor impact on CH4 uptake in the growing season. Of all the environmental variables included, soil temperature and moisture were the main drivers of GHG flux variation. The sustained flux global warming potential from the PV arrays significantly increased by 8.14% compared to the ambient grassland. Our evaluation models identified that the GHG footprint of PV arrays during the operation period on grasslands was 20.62 g CO2-eq kW h-1. Compared with our model estimates, GHG footprint estimates reported in previous studies were generally less by 25.46-50.76%. The contribution of PV power generation to GHG reduction may be overestimated without considering the impact of PV arrays on hosting ecosystems.
Keywords: carbon footprint; greenhouse gas; microclimate; renewable energy; solar parks.