Surface Temperature Measurement from Space: A Case Study in the South Western Cape of South Africa

Abstract

This paper reports on the use of Landsat Thematic Mapper (TM) thermal infrared (TIR) and Transformed Normalised Difference Vegetation Index (TNDVI) data to map summer surface temperature distribution in a portion of the South Western Cape (South Africa). It adopts a ground truth calibration approach, as a possible solution for alleviating the paucity of spatial temperature data. Means and standard deviations of the TIR and transformed NDVI of all pixels within a radius of 120 m around each ground truth point were regressed on spatially collocated surface-observed temperature data. Linear multiple regression analysis showed that mean TIR and standard deviation of TNDVI were effective in accounting for 43 % of variation in surface temperature. The regression equation was used to generate a temperature map at a spatial resolution of 30 m, showing good agreement with expected mesoscale spatial temperature patterns in terms of general climatic principles. The temperature map provides a useful tool for depicting and assessing the spatial variation in daily temperatures, providing information currently unavailable to the fruit-producing farmers in the South Western Cape.