Long-term analysis of the urban heat island effect using multisource Landsat images considering inter-class differences in land surface temperature products
Xiong Xu,Haoyang Pei,Chao Wang,Qingyu Xu,Huan Xie,Yanmin Jin,Yongjiu Feng,Xiaohua Tong,Changjiang Xiao
DOI: https://doi.org/10.1016/j.scitotenv.2022.159777
2023-02-01
Abstract:It is imperative to quantitatively analyze the long-term temporal and spatial characteristics of the urban heat island (UHI) effect on cities for applications, such as urban expansion and environmental protection. Owing to the high spatial resolution and availability of long time-series data, remote sensing images from Landsat satellites are widely used for land surface temperature (LST) retrieval. However, limited by the satellite revisit cycle and image quality, the use of multisource Landsat images in a long-term study of the UHI effect is inevitable. Nonetheless, owing to the differences among multisource sensors, such as Landsat-7 and Landsat-8, there may be apparent deviations in the LST results retrieved from different sensor data, which are obtained from the same area and under similar circumstances. Consequently, it is necessary to build a relationship between the LST results generated from multisource Landsat sensors for future research on the UHI effect.In this study, Shenzhen city was studied to explore the fitting relationship between the corresponding LST products from Landsat-7 and Landsat-8 images obtained from adjacent dates with similar climatic conditions. Furthermore, factors affecting the fitting models, such as land cover types, seasonal and inter-annual differences, were analyzed. The constructed fitting model had a strong relationship with land cover types but a relatively weak relationship with seasonal and inter-annual differences; this indicates that a pseudo Landsat-8-based LST product can be generated from a Landsat-7-based LST product using a model fitted by a Landsat-7/8 pair obtained from adjacent years (or different seasons). Finally, by considering the consistency between LST products from multisource Landsat images, the spatiotemporal variations in the UHI effect in Shenzhen can be accurately explored using long time-series data.
environmental sciences