Impact of 3-D structures on directional effective emissivity in urban areas based on DART model

Thermal infrared (TIR) remote sensing is effective to monitor urban land surface temperature (LST) on a large scale. The accuracy of the inversion of LST depends strongly on the emissivity. The emissivity is complex in urban areas and is largely dependent on three-dimensional (3-D) urban structure and different building materials, also called effective emissivity. The urban canyons have homogeneous temperatures to avoid the difficulty of defining the emissivity of non-isothermal surfaces. In this paper, we use the 3-D discrete anisotropic radiative transfer (DART) model to simulate the emission and propagation of thermal radiation in urban scenes by a flux tracking method and generate remote sensing images. The impact of 3-D structures on the effective emissivity was analyzed and simple relations between effective emissivity and 3-D structure parameters were proposed. Results show that the increase of emissivity at nadir is well related to wall frontal area index (F) and the mean sky view factor of flat surfaces (i.e., roof and road, SVFweighted), whereas emissivity directionality is well related to road's sky view factor (SVFroad). Overall, emissivity directionality (reaches 0.069 if reflectance is 0.2 for the road and 0.1 for the other components) due to material mixture and nadir emissivity increment (increase by 0.0646 for material reflectance R of 0.1) caused by 3-D structures can be larger than other emissivity sources. This study helps to understand the impact of 3-D structures and different material emissivity on urban effective emissivity and offers insights for addressing urban emissivity directionality in multi-angular remote sensing missions.