A comparison of modeled and observed urban surface temperatures in Toulouse, France

The ability of the Town Energy Balance Model (TEB, Masson 2000) to model surface temperatures in an urban canyon is evaluated using observations from the CAPITOUL field campaign conducted in Toulouse France. The study area in central Toulouse incorporates six walls, three roads, and one roof surface that were monitored with infrared radiometers. To compare observed surface temperatures to the TEB output, a weighted averaging scheme is developed for the wall and road surfaces in order to produce one observed temperature representative of each surface. Two methods are proposed to correct for the effect of traffic influencing the road temperature readings. A study period comprised of three 30-day sub-periods (representing summer, fall, and winter) allows for the performance of TEB to be assessed for each surface relative to season, time of day, and meteorological conditions. Case studies are used to further emphasize model performance. Overall, TEB-modelled surface temperatures are concluded to be robust. TEB is found to be most accurate when modelling wall temperatures. Road temperatures are reproduced well but with less accuracy than the walls. The model performs relatively poorly when modelling roof temperatures but the comparison is limited by the number of roofs measured. Observed and modelled temperatures compare better during nighttime hours and during the fall and winter. Maximum differences occur with high solar radiation and high surface temperatures when sensitivity to surface characteristics and atmospheric conditions are greatest.