Research on the spatiotemporal coupling relationships between land use/land cover compositions or patterns and the surface urban heat island effect

Urbanization leads to changes in landscape configuration and land use/land cover (LULC) patterns, and these changes are important factors affecting the surface urban heat island (SUHI) effect. However, from the perspective of spatiotemporal changes, quantitative analytical results regarding the impacts of the LULC composition, configuration, and pattern in inland plateau lakeside cities on the SUHI effect, and the responsive relationships among these factors remain unclear. By combining satellite remote sensing data with analytical methods, such as urban–rural gradients, spatial statistics, and landscape pattern indices, the impacts of LULC changes on the SUHI effect in Kunming, China, are revealed. The results show the following. (1) The explosive growth in impervious surfaces (ISs) caused by urbanization, leading to changes in the LULC composition, configuration and pattern, is the main reason for the deterioration of the SUHI effect. Over the past 30 years, Kunming’s ISs have increased by 304.58 km², SUHI has expanded by 764.26 km², and the regional average land surface temperature (LST) has increased by 1 °C. (2) This study also found that a large area of bare ground is another important reason for the sharp rise in LST, explaining why bare land (BL) has the highest average LST (28.72 °C). (3) The pattern of LULC can well explain the spatial distribution characteristics of SUHIs. The normalized difference built-up index (NDBI), normalized difference bareness index (NDBaI), and LST have the same change curve along the urban–rural gradient, while the normalized difference vegetation index (NDVI), modified normalized difference water index (MNDWI), and LST have opposite trends. (4) ISs and water body (WB) are the main types of warming and cooling, respectively, but the warming effect of ISs is greater than the cooling effect of WB. From the average value of the correlation coefficient with LST, NDBI (0.84) > MNDWI (-0.63). (5) Kunming’s remote sensing index values do not have simple linear relationships with the LST. NDBaI, NDBI, and LST show significant exponential relationships, and NDVI, MNDWI, and LST show significant quadratic polynomial relationships. (6) The dominant landscape type determines the correlation between the landscape shape index (LSI) and the LST of green spaces (GSs). (7) Adopting a simple and regular landscape layout can effectively reduce the SUHI effect. These research results could provide a scientific decision-making basis for the spatial urban planning and ecological construction of Kunming and could have practical significance for guiding the green, healthy, and sustainable development of the city.