Abstract: Photovoltaic power generation plays a crucial role in the structural transformation of China"s power system. Investigating the spatiotemporal characteristics and influencing factors of spring solar radiation variability in the Southern Power Grid region is highly significant for predicting its photovoltaic generation potential and optimizing power dispatch. Utilizing surface solar radiation data, sea surface temperature, and atmospheric circulation data from 1979 to 2023, methods including Empirical Orthogonal Function decomposition, Mann-Kendall mutation test, correlation analysis, and composite analysis were employed to examine the spatiotemporal features of spring surface solar radiation changes in this region and their influencing factors. The results indicate an overall increasing trend in radiation over the study area during the past four decades, with the most significant increase observed in spring. Further investigation revealed that the spatial distribution of spring radiation variability is dominated by two primary modes: an east-west dipole pattern and a regionally uniform pattern. Additionally, it was found that winter sea surface temperature conditions in key oceanic regions possess predictive potential for spring photovoltaic generation, serving as an effective precursor signal for cross-seasonal forecasting.