Thermal Infrared Radiative Transfer Modeling in Urban Areas by Considering 3-D Structures and Sunlit–Shadow Temperature Contrast

Land surface temperature (LST) is a crucial parameter needed to study the thermal environment in urban areas. Currently, it can be restored from thermal infrared (TIR) measurements based on various LST retrieval algorithms. However, the expected urban LST retrieval accuracy of < 1 K is difficult to achieve because knowledge is lacking on how to correct the impact from the surface of 3-D structures and the sunlit–shadow temperature contrast. Although an analytical TIR radiative transfer model over urban area (ATIMOU) has been proposed, the temperature contrast between sunlit and shadowed areas has not been well-managed yet, thus leading to its inapplicability in daytime TIR observations. This study develops an extended ATIMOU (E_ATIMOU) that considers the impact from both 3-D structures and sunlit–shadow temperature contrast. According to the simulations based on E_ATIMOU, if such impact is not properly accounted for, a 4.43-K bias can be potentially introduced to the ground brightness temperature of a street canyon under the condition of wavelength of
$10~\mu \text{m}$
, ratio “sunlit-road area/total-road area” of 0.5, shadowed wall and road temperature of 300 K, and the sunlit–shadow temperature contrast of 5 K, which emphasizes the necessity of addressing this impact during the LST retrieval in urban areas. Moreover, E_ATIMOU has also been validated by intercomparing with the discrete anisotropic radiative transfer (DART) model. The discrepancy between the two models for the calculated ground brightness temperatures is found to be < 0.1 K for various urban scenarios, indicating that the E_ATIMOU is in good agreement with DART.