Land surface temperature (LST) is increasingly important for various studies assessing land surface conditions, such as studies of urban climate, evaporation and vegetation pressure. The Landsat series of satellites has the potential to provide LST estimates with high spatial resolution, which is particularly suitable for local or small-scale studies. Many studies have proposed LST retrieval algorithms for the Landsat series, and some datasets are available online. However, these datasets usually require users to be able to process large amounts of data. Google Earth Engine (GEE) is an online platform designed to allow remote sensing users to easily perform big data analysis without increasing demands on local computing resources. However, the high spatial resolution LST dataset is not currently available in GEE. Here we provide a code library to compute LSTs from Landsat 4, 5, 7 and 8 in GEE. This code is freely available to users to compute Landsat LST as part of any analysis in GEE. Preface – Artificial Intelligence Tutorial
The first step in LST production is to match the TCVW estimates from the NCEP reanalysis data to the Landsat observation times. However, due to the coarse resolution of the model, TCWV values are nearly constant in the selected area and, therefore, are not shown. The TIR surface radiance required to calculate the LST comes from the ASTER GEDv3 dataset. Figure 4b shows the corresponding emissivity map for ASTER band 14, which is closest to the Landsat TIR band. ASTER NDVI values were used to calculate the corresponding FVC, as shown in Fig. 4c. FVC and emissivity values are fairly uniform over selected regions. FVC is generally quite low, with values between 0.3 and 0.5 in most areas. The resulting bare-ground radiance map shows values of approximately 0.96–0.97, which is consistent with values typically found in spectral libraries for the 11 μm region [53