Coordination Analysis of Temporal and Spatial Variation of Land Surface Temperature and Urban Development in Shihezi Oasis
- ZHANG Xue-ling, KASIMU·Alimujiang, LIANG Hong-wu
Related Articles |
As an important place for human habitation and prosperity development in arid and semi-arid regions, the oasis's unique "cold island effect" characteristics have profound impact on the quality of the environment for human living. Based on Landsat remote sensing images, the cold island index was calculated by inverse calculation for Shihezi City's land surface temperature (LST) in 2006, 2011, 2017, and 2021 using the radiation transfer equation model, and the spatial autocorrelation of LST was then analyzed. After extracting land use information by adopting random forest classification, based on the Geodetector model, nine types of data, including five remote sensing indicators, NDVI, NDBI, MNDWI, SI, and NDMI, and land use, DEM, slope, and night lighting data, were selected to detect the spatial heterogeneity of LST in Shihezi City. The degree of coordination between town development and LST was finally determined by using the coordination model. The outcomes of the study show that, the construction land increased in all directions with the total area growing from 83.42 km2 to 164.05 km2 within 15 years; forest land increased gradually, while cultivated land, grassland, and water areas decreased year by year. The spatial distribution of LST is strongly influenced by land use type. The areas of extremely high temperature, high temperature, and medium temperature zones increased with the expansion of the city. The cold island index of the four periods, was 0.53, 0.47, 0.39, and 0.35, respectively, and the cold island intensity reached the highest decline rate since 2017 and slowed in 2021. The examination of the spatial autocorrelation of LST reveals that LST had a strong positive and significant spatial correlation . The results of factor detection show that the q value of NDBI was the highest, indicating that town development had a strong influence on LST; while the interaction detection results show that the influence of the two-factor interaction on LST was greater than that of the single factor, with the explanatory power of DEM∩NDMI and DEM∩NDBI on LST being the highest. Finally, in the early stage of town development, LST changes ahead of town development, but in the late stage of town development, LST and town development were restricted in some extent, and LST changes laged behind the town development.