Evaluation of different WRF-Urban Canopy Models performance in city-scale numerical simulations in Tehran Metropolis

Simulation of near-surface weather parameters is a challenging process, especially in urban areas, because it is difficult to precisely identify surface characteristics in urban micro-scales. Different urban parameterizations for the representation of urban structure are coupled with numerical weather or climate models to improve the accuracy of the micro-scale simulations. In this study, the numerical results of the Weather Research and Forecasting model (WRF) with three different urban configurations, namely no urban canopy or the SLAB scheme, Single-Layer Urban Canopy Model (SLUCM) and Multi-Layer UCM or the Building Effect Parameterization (BEP) in the simulation of near-surface air temperature, relative humidity and wind speed are evaluated against the observations in Tehran Metropolis, during 15 to 29 June 2016. Overall, results show that SLUCM and BEP predict meteorological parameters more accurately than SLAB scheme. Although the performance of the model is not the same in different weather stations, comparing SLUCM and BEP results, on average, over four stations of Tehran shows that BEP results in minimum errors and the maximum Pearson coefficients. In addition, the more intense night-time urban heat island is also simulated in BEP (over 2.5°C) in comparison to SLUCM (1.5°C) and SLAB (0.5°C). However, the daytime UHI intensity is approximately simulated with the same intensity in the three mentioned simulations. Since high-resolution numerical simulations are time-consuming and expensive, current results can be used in other related studies to avoid extra costs.