Assessment of thermal anisotropy on remote estimation of urban thermal inertia

Abstract

Thermal inertia over vast earth surfaces is a crucial parameter in many related disciplines. However, remote estimates of urban thermal inertia show anisotropy effects due to thermal anisotropy. This study investigates the impacts of thermal anisotropy on the estimation of urban thermal inertia. We present the concepts of DTI (directional thermal inertia) and DATI (directional apparent thermal inertia) to describe this anisotropic effect. A combined approach to estimating thermal inertia named NLS (nonlinear least square) is proposed as a compromise solution using temporal temperature measurements. Intercomparisons between methods are indirectly conducted by predicting surface temperatures, and the results indicate the NLS has higher accuracy of 1 to 2K. The DTI estimation over an urban scale model, together with flat concrete and grass surfaces, reveals that the DTI intensity is significant when DRTs (directional radiometric temperatures) are used as inputs. Over the scale model, the DTI and DATI values range from 0.028 to 0.038K−1 and from 1530 to 2970W∙s1/2∙m−2∙K−1 if the zenith and azimuth are between −60° and +60° and between 0° and +360°, respectively. Further discussions demonstrate that DRT intensity has an approximate linear relationship with DTI intensity and that it thus could be used as a predictor of DTI intensity. We finally propose that using complete urban surface temperatures instead of the DRT would be better to estimate urban thermal inertia from the perspective of surface energy balance.