Longwave Thermal Infrared Atmospheric Compensation Using In Situ Scene Elements: Accurate Temperature Determination of Blackbody and Graybody Targets From Small Unmanned Aircraft Systems

Accurate determination of surface temperature from small unmanned aircraft system (sUAS) acquired longwave infrared (LWIR) imagery requires compensation for atmospheric transmission, upwelling path radiance, and reflected downwelling radiance. The multiple-altitude technique has previously been demonstrated as an effective approach for estimating atmospheric transmission and upwelling radiance, but its practical application has relied on the assumption of high-emissivity targets of interest for which reflected downwelling radiance is negligible. This assumption limits applicability in agricultural and natural-scene environments, where canopies, soils, senescent vegetation, and man-made reference materials often behave as graybodies.

This study extends the multiple-altitude framework by incorporating a multiple-angle observation strategy for estimating downwelling radiance. LWIR imagery was collected over ground reference blackbody and graybody targets using a radiometrically calibrated sUAS at typical sUAS altitudes. Retrieved target temperatures were validated against independent \insitu measurements and evaluated using absolute error metrics and statistical significance testing.

Results show that the multiple-altitude technique remains effective for high-emissivity targets, but produces systematic temperature overestimation for lower-emissivity graybody surfaces when reflected downwelling radiance is neglected. Incorporating the multiple-angle compensation reduced average graybody temperature error from approximately 2.76K to 1.16K across the analyzed flights. These results demonstrate that explicit treatment of reflected downwelling radiance is necessary for physically consistent temperature retrieval from realistic sUAS thermal imagery and is directly relevant to agricultural applications requiring accurate canopy or soil temperature, including crop water stress assessment, evapotranspiration estimation, irrigation scheduling, and phenotyping.