Simulation and prediction of urban heat island effect driven by land use change: a case of Changsha-Zhuzhou-Xiangtan region

Abstract

With the acceleration of urbanization, the rapid expansion of urban land use has led to an intensification and expansion of the urban heat island effect. This study focuses on the Changsha-Zhuzhou-Xiangtan region as the study area. Through geographical detectors, it analyzes the contribution intensity of land cover type data, DEM, GDP, and population density to the urban heat island. Using the CA-Markov model, it conducts a predictive analysis of land use conditions in the Changsha-Zhuzhou-Xiangtan region in 2025. Furthermore, based on the ANN-CA model, it predicts the intensity of the urban heat island in the Changsha-Zhuzhou-Xiangtan region for both 2025 and 2040. The results indicate that the proportion of construction land area, GDP, and DEM are the most influential factors contributing to the urban heat island. The prediction accuracy of the ANN-CA model for the urban heat island in 2015 and 2020, starting from 2000 and 2005, reached 86.12% and 94.8%, respectively, demonstrating the reliability of the ANN-CA model in predicting the urban heat island. Compared to 2020, the combined area proportion of strong and hot urban heat island regions in the Changsha-Zhuzhou-Xiangtan region increased by 1.04% in 2025 and 1.02% in 2040. During the period from 2020 to 2040, the heat island effect in the Changsha-Zhuzhou-Xiangtan region showed a slow increase, indicating that the heat island effect will be controlled to some extent overall.