Abstract: With the acceleration of global urbanization, the urban heat island (UHI) effect has posed serious risks to the functioning of urban ecosystems. However, researchers have focused predominantly on analyzing the impacts of land use patterns on UHI dynamics, offering limited actionable guidance for targeted environmental management. To address this shortcoming, by taking the central Guizhou urban agglomeration as the study area, morphological spatial pattern analysis (MSPA), circuit theory, and the complex network topology research method were integrated to identify and remediate the thermal environment network through different edge-deleting methods. In addition, the remediation effect was verified via robustness analysis. The results indicated the total area of heat island patches expanded from 7484.63 km² in 2010 to 11 096.94 km² in 2022, with core-type patches becoming dominant. Both the number of sources and corridors in the thermal environment network and the complexity of the structure increased. Compared with the corridor remediation effect, the high-degree-first (HDF) strategy performed better in the robustness analysis, with the lowest original degree of connection robustness (decreasing from 0.82 to 0.73) and the fastest rate of decrease. Moreover, the node restore robustness reduction effect was the best, and the edge restore robustness reduction effect was comparable to that of the high-betweenness-first (HBF) strategy.This study could provide actionable insights into the mitigation of the risk of the urban thermal environment and the promotion of sustainable urban development patterns.