Assessing the dynamic relationship among land use pattern and land surface temperature: A spatial regression approach

Abstract

ABSTRACT Drastic changes in land use pattern in recent times has translated into severe alteration of urban surface temperatures, whereby the increase in impervious surfaces results in modification of the thermal structure of the city and consequently leads to the problem of Urban Heat Island. Conversely, creation of green surface within the city reduces the temperature and forms a cooling effect. The present study investigates the impact of the shape, pattern, and configuration of the two dominant land use/land cover classes (built-up and green surface) on the Land Surface Temperature (LST) of Hyderabad city. Different class-level landscape metrics at neighborhood level have been employed to evaluate the spatial variation of LST. Significant Moran’s I cluster value indicates spatial dependencies of landscape metrics and LST. The local Geographical Weightage Regression (GWR) model has also been executed to examine the relative importance of land use pattern on LST. Results reveal that during 2002 and 2015, the increasing fragmentation of built-up and LSI have positively influenced the LST, whereas, urban green surface, the PLAND, and IJI have negatively affected the LST. The estimated LST is found to be highly sensitive to the configuration and changes in land-use pattern. The spatial regression result suggests that with the expansion of built-up in the neighborhood, there is a discernible increase in the LST. Similarly, fragmented green cover reduces the neighboring temperature. The outcome of this study may enable planners to obtain insight into where urban management and green surface planning is necessary for mitigating the surface temperature.