An Algorithm for Separating Soil and Vegetation Temperatures With Sensors Featuring a Single Thermal Channel

Soil and vegetation temperature separation (SVTS) is a crucial process in various fields, such as the study of evapotranspiration. An operational and novel algorithm for separating soil and vegetation temperatures from sensors that feature a single thermal channel was developed to analyze high-heterogeneity croplands. The a priori knowledge on interrelationships among neighboring pixels was coupled to both the radiation transfer equation and the conceptual thermal anisotropic model to increase the solvability of forward anisotropic models through the Bayesian theorem. Model sensitivity analysis suggests that component fractions and reference temperatures are the two main factors that control the accuracies of the inversion results. Some validation options, which include the air temperature data from local weather stations, computer simulations, up-scaling techniques, and intercomparisons among different approaches, were selected as indirect techniques in verifying the inverted results. These results demonstrated that the proposed inversion technique reached an acceptable level of accuracy and stability, which highlights the practicalities of monowindow thermal sensors in the SVTS.