Abstract

Land surface reflectance and temperatures data products are generated from data acquired with the Operational Land Imager and Thermal Infrared Sensor on the Landsat 8 and 9 satellites. The surface temperatures (ST) in the Level 2 surface products (L2SP) are produced using a single channel algorithm, which takes no consideration of the presence of cold thin cirrus clouds, that re-emit thermal radiation at lower temperatures. The errors in the ST data products over cirrus-contaminated areas can be larger than 10 K. Our recent analysis of Level 1 top of the atmosphere (TOA) data products (L1TP) and L2SP data products acquired over different geographic regions, show that it is possible to use the correlations between the Level 1 TOA reflectances of Band 9 (referred as the cirrus band) and the Level 2 ST data products to remove thin cirrus induced errors in the daytime Level 2 ST data products over water. The corrected daytime ST datasets have standard deviations reduced to about 1 K, similar to results with no cirrus present. End users of the Level 2 ST data products are encouraged to test this method to make cirrus corrections to the existing daytime ST data products, and to more broadly apply its results and further its applicability.

IPC Classification

Keywords