A study of the urban boundary layer using different urban parameterizations and high-resolution urban canopy parameters with WRF

In the last two decades, mesoscale models (MMs) with urban canopy parameterizations have been widely used to study urban boundary layer processes. Different studies show that such parameterizations are sensitive to the urban canopy parameters (UCPs) that define the urban morphology. At the same time, high-resolution UCP databases are becoming available for several cities. Studies are then needed to determine, for a specific application of an MM, the optimum degree of complexity of the urban canopy parameterizations and the resolution and details necessary in the UCP datasets. In this work, and in an attempt to answer the previous issues, four urban canopy schemes, with different degrees of complexity, have been used with the Weather Research and Forecasting (WRF) model to simulate the planetary boundary layer over the city of Houston, Texas, for two days in August 2000. For the UCP two approaches have been considered: one based on three urban classes derived from the National Land Cover Data of the U.S. Geological Survey and one based on the highly detailed National Urban Database and Access Portal Tool (NUDAPT) dataset with a spatial resolution of 1 km². Two-meter air temperature and surface wind speed have been used in the evaluation. The statistical analysis shows a tendency to overestimate the air temperatures by the simple bulk scheme and underestimate the air temperatures by the more detailed urban canopy parameterizations. Similarly, the bulk and single-layer schemes tend to overestimate the wind speed while the multilayer schemes underestimate it. The three-dimensional analysis of the meteorological fields revealed a possible impact (to be verified against measurements) of both the urban schemes and the UCP on cloud prediction. Moreover, the impact of air conditioning systems on the air temperature and their energy consumption has been evaluated with the most developed urban scheme for the two simulated days. During the night, this anthropogenic heat was responsible for an increase in the air temperature of up to 2°C in the densest urban areas, and the estimated energy consumption was of the same magnitude as energy consumption obtained with different methods when the most detailed UCP database was used. On the basis of the results for the present case study, one can conclude that if the purpose of the simulation requires only an estimate of the 2-m temperature a simple bulk scheme is sufficient but if the purpose of the simulation is an evaluation of an urban heat island mitigation strategy or the evaluation of the energy consumption due to air conditioning at city scale, it is necessary to use a complex urban canopy scheme and a detailed UCP.