The correlation analysis between meteorological factors and power load is critical to power load forecasting, and it is necessary to correct the empirical model of correlation according to actual data. Based on the comprehensive meteorological index, accumulated temperature effect, and the empirical formula of correlation analysis between power load and meteorological factors, a method for correcting the correlation model of meteorological load in summer is proposed. Load trend analysis and Python crawling are used to extract meteorological load and meteorological data to improve the accuracy of analytical data. By comparing and analyzing the correlation coefficients between load and single meteorological factors, comprehensive meteorological indices, and two kinds of accumulated temperature effect corrections, combined with the coincidence degree of load and meteorological indicators over time, the optimal index parameters suitable for correlation analysis are determined. And then the fitting relationship between meteorological load and optimal index parameters is constructed. The proposed method is applied and verified by taking the main urban area of Beijing in summer 2019 as an example. The results show that compared with single meteorological factors, there is a stronger correlation between power load, meteorological load, and comprehensive meteorological index. Among the comprehensive meteorological indices, the heat index based on daily average temperature has the highest correlation coefficient with meteorological load. Among the two accumulated temperature correction methods, the method considering the cumulative effect coefficient has a better correction effect on temperature, and the correlation coefficient between the corrected temperature and meteorological load increases by 7.39%. The correlation between the heat index based on the corrected temperature and meteorological load is higher than that before correction, which is closer to the changing trend of load. The coincidence degree between the fitting relationship of meteorological load constructed with heat index and corrected temperature as independent variables and the actual value is higher than that of the referenced empirical formula.