Lake surface water temperature (LSWT) is an important factor in aquatic environment, which directly affects watershed ecosystem and biodiversity. Precise LSWT measurement and prediction is essential to control and improve the aquatic ecological environment of the river basin, and is also the key to prevent and control the outbreak of cyanobacteria bloom. Focusing on Dianchi Lake, 54 water quality parameters (LSWT, chlorophyll a, pH, permanganate index, dissolved oxygen, etc.) of 10 water quality monitoring sites from 2005 to 2016 are used as the data set. A hybrid forecasting model is presented composed of εsupport vector regression (εSVR), principal component analysis (PCA) and back propagation artificial neural network (BPANN). Moreover, Kriging method is combined with geographic information system (GIS) to realize the scene reproduction of the historical changes of the Dianchi lake LSWT and water quality in the past 12 years and the trend simulation of the next 5 years. Results show that the average relative error of the model is 0.5%, the mean square error is 1.452 3, R2 is 0.904 9. Spatial visualization results indicate that the region with LSWT over 20 ℃ diffuses obviously from north to south. The outbreak of cyanobacteria bloom changes from locally to globally, which is related to the expansion of Kunming urbanization and meteorological environment.