The spatiotemporal changes of urban thermal environment influenced by landscape patterns—based on panel data model

Abstract With the development of urbanization, the spatial and temporal distribution characteristics of the urban landscape pattern play a decisive role in the intensity of the urban heat island (UHI) effect. Most of the current dynamic studies utilizing single cross-section data are labor-intensive and lack continuity. A panel data model was constructed to study the relationship between the urban heat island effect and landscape pattern in Kunming from 1995 to 2020 at four different scales. An ordinary least squares model was constructed to verify the reliability of the panel data model results. The results indicate: (1) Kunming City's landscape pattern has undergone significant changes over time, with an increase in artificial patches and a decrease in natural patches; (2) The UHI effect in Kunming City is evident, with the low-temperature zone gradually decreasing to disappearance and a significant increase in the high-temperature zone; (3) With an increase in scale, the number of influencing variables continues to grow, and the impact of artificial patches gradually intensifies;(4) The normalized difference water index (NDWI) has the greatest influence on the surface thermal radiance, and the cooling effect is more obvious with the increase in scale. Unlike previous studies, the spatial configuration of the landscape in Kunming City had a stronger effect on the UHI effect than the landscape grouping and vegetation index;(5) This paper introduces the panel data model into the discussion of the UHI for the first time, providing a new method for better understanding the changing patterns of the urban thermal environment.