Spatial and temporal variations in the relationship between lake water surface temperatures and water quality - A case study of Dianchi Lake

Global warming and rapid urbanization in China have caused a series of ecological problems. One consequence has involved the degradation of lake water environments. Lake surface water temperatures (LSWTs) significantly shape water ecological environments and are highly correlated with the watershed ecosystem features and biodiversity levels. Analysing and predicting spatiotemporal changes in LSWT and exploring the corresponding impacts on water quality is essential for controlling and improving the ecological water environment of watersheds. In this study, Dianchi Lake was examined through an analysis of 54 water quality indicators from 10 water quality monitoring sites from 2005 to 2016. Support vector regression (SVR), Principal Component Analysis (PCA) and Back Propagation Artificial Neural Network (BPANN) methods were applied to form a hybrid forecasting model. A geospatial analysis was conducted to observe historical LSWTs and water quality changes for Dianchi Lake from 2005 to 2016. Based on the constructed model, LSWTs and changes in water quality were simulated for 2017 to 2020. The relationship between LSWTs and water quality thresholds was studied. The results show limited errors and highly generalized levels of predictive performance. In addition, a spatial visualization analysis shows that from 2005 to 2020, the chlorophyll-a (Chia), chemical oxygen demand (COD) and total nitrogen (TN) diffused from north to south and that ammonia nitrogen (NH3-N) and total phosphorus (TP) levels are increases in the northern part of Dianchi Lake, where the LSWT levels exceed 17 degrees C. The LSWT threshold is 17.6-18.53 degrees C, which falls within the threshold for nutritional water quality, but COD and TN levels fall below V dass water quality standards. Transparency (Trans), COD, biochemical oxygen demand (BOD) and Chla levels present a close relationship with LSWT, and LSWTs are found to fundamentally affect lake cyanobacterial blooms. (C) 2017 Elsevier B.V. All rights reserved.