Annual and diurnal temperature cycle modelling of a merged multi-annual ECOSTRESS and Landsat land surface temperature dataset

Annual and diurnal temperature cycle (ATC and DTC) models are proven tools to estimate LST dynamics. Based on solar geometry, they typically require 3 to 6 parameters determined by fitting them to a LST dataset. The combined ADTC model employed here uses five controlling parameters to represent both cycles simultaneously, i.e., it allows to estimate the average annual and diurnal temperature cycle dynamics from five fitted parameters. High spatial resolution (70 m) model parameters were obtained for first time from a six year (2018 – 2023) datacube of merged ECOSTRESS and Landsat LSTs. Since LSTs at actual observation times were modelled, the variable overpass time of ECOSTRESS helped to sample the diurnal cycle. All available ECOSTRESS and Landsat LSTs over the four European cities Karlsruhe (Germany), Paris (France), Madrid (Spain) and Valencia (Spain) were obtained. The mean RMSE of the modelled LST over these study areas was 3.8 K. The high-resolution model parameters were compared against those obtained for MODIS for the same period and areas. The results showed RMSEs of 1.6 K, 2.0 K, 6.5 K, 0.2 h and 6.8 days for annual minimum temperature, annual amplitude, daily maximum amplitude, mean thermal noon and lag with respect to summer solstice, respectively. ... mehrThe larger RMSEs observed for maximum daily amplitude and lag with respect to summer solstice are attributed to insufficient data around thermal noon in the high-resolution dataset. The parameters were analyzed over rural and urban areas and show clear potential for more detailed urban thermal studies and land cover change monitoring.