Search

Abstract: The measurements of surface temperature are prone to important directional anisotropy in relation to the structure of the urban canopy and the radiative and energy exchanges inside it. Following the work of Lagouarde et al. 2010 that describes the daytime conditions, this paper focuses on the experimental analysis and modelling of the nighttime directional anisotropy. An extensive data set of airborne thermal infrared (TIR) measurements was collected over the city of Toulouse, France in the framework of the CAPITOUL project (Masson et al., 2008). The TIR measurements use a pair of thermal cameras equipped with wide angle lenses installed aboard a small aircraft. Three flights were made between 21:45 and 23:15 UTC, one in autumn and 2 in winter during 2004 and 2005 intensive operation periods (IOPs). The experimental results show that (i) the nighttime TIR directional anisotropy remains lower than 1°C for zenithal view angles up to 50°, and (ii) the nighttime anisotropy is insensitive to the azimuthal viewing direction. A modelling approach that combines a simplified 3D representation of the urban canopy with 2 energy transfer models, TEB and SOLENE, is then proposed. It confirms the experimental results quite well. Possible uncertainties on the geometry of the canopy and on most important thermo-radiative surface parameters are shown to have only little impact on the modelled TIR directional anisotropy. The time evolution of TIR anisotropy throughout night is simulated: for the fall and winter conditions studied the effects of thermal inertia completely vanish about 3 to 4h after sunset, and the anisotropy never exceeds 2°C up to 60° zenithal view angles. [Copyright &y& Elsevier]

Abstract: Measurements of thermal infrared (TIR) directional anisotropy (difference between off-nadir and nadir brightness temperatures) performed over the city of Toulouse using a method based on the use of 2 airborne TIR cameras are presented. Results from 3 flights at different times during a summer day (July 15th 2004) and from 1 flight in winter (February 25th 2005) all confirm important anisotropy (up to 10°C) and hot spot effects as previously reported in literature. A simple simulation approach is then proposed. It is based on the aggregation in any viewing direction of 6 directional temperatures (sunlit/shaded walls/streets/roofs) weighted by their corresponding surface ratios within the scene viewed. The city is described by 18 canyon streets oriented in all directions by 10° steps and the 6 temperatures are determined by integrating simulations of the energy balance model SOLENE repeated for the 18 canyon streets. The surface ratios are computed from images of the studied area generated with the POV-Ray software (Persistence of Vison Raytracer, http://www.povray.org/). This method is described in detail. The modelled anisotropy compares favourably with the measurements on all dates, despite a systematic underestimation ranging between 15 and 30%. The possible sources of discrepancy including sensitivity to the aspect ratio and to the surface parameters and possible impact of microscale structures are briefly discussed and several improvements of the modelling system are suggested. [Copyright &y& Elsevier]