Abstract: The power load obviously increases in summer due to significant increase of air temperature. In allusion to the difficulty of obtaining meteorological loads and the low forecasting accuracy of meteorological load, a new method to forecast meteorological load was proposed. Firstly, to obtain accurate data of meteorological load, the growth curve was applied to describe the growth characteristics of baseload and by means of eliminating baseload the data of meteorological load could be obtained. Secondly, considering the temperature accumulation effect of high-temperature weather in summer, daily highest temperature in the high-temperature weather had to be revised, thus a meteorological load based temperature correction method and corresponding model were proposed. Finally, a particle swarm optimization-extreme learning machine load forecasting model was established to forecast the total load and meteorological load. Analysis results of numerical example show that based on the growth curve and temperature accumulation effect the load forecasting results are improved, and the effectiveness of the proposed algorithm and model are verified.