A reconstruction dataset of land surface temperature at high spatial and temporal resolutions in Chengdu city, China from 2001 to 2022

The dataset is composed of four folders, and each folder is divided into 21 subfolders from 2000-2022. The data content and organization of each folder are as follows:(1) Folder 1 contains the different spatial resolution raw thermal infrared image applied to spatiotemporal fusion in study area. The medium spatial resolution satellite data includes Landsat TM, Landsat 8, Landsat 9, HJ-1B CCD, HJ-1B IRS. The folder size is 77.2 GB.(2) Folder 2 contains downscaling land surface temperature dataset at 30 m using machine learning algorithm. The folder size is 724 MB.(3) Folder 3 contains land surface temperature dataset with low spatial resolution. The folder size is 30.1 MB.(4) Folder 4 contains high spatiotemporal land surface temperature dataset by spatiotemporal fusion method in the study area. The folder size is 14.0 GB.The dataset totals 91.9 GB.