The impact of historic (1973, 1985, 1993, 1998 and 2005) land use/cover on near-surface meteorological variables was investigated for the arid Phoenix (Arizona, USA) metropolitan area using the Weather Research and Forecasting Model (WRF). In order to account for urban thermal and dynamic effects on the atmosphere an urban canopy model (UCM) was applied with WRF. Modifications were made to the WRF UCM code in order to account for specific conditions in the Phoenix region including adjustments of anthropogenic heating, soil moisture due to landscaping irrigation, urban vegetation classes and physical parameters of urban land use classes.

Observations from surface stations and simulated results both indicated an intensification and expansion of the urban heat island with urban growth during the three decades (1973-2005) and a continuation of this trend for simulated potential future urban development. Significant spatial variations in near-surface air temperature across the metropolitan region were detected that depended mainly on the distance from the urban core, levels of irrigation for urban landscaping, and building characteristics. The night-time urban heat island was strongest during early summer and fall when synoptic conditions are usually week and low levels of atmospheric moisture dominate the climatology. An afternoon urban heat island was found during the monsoon season in July and August. In the model this is in part due to the decreased vapor pressure deficit and the subsequent reduction in latent heat fluxes from landscaping transpiration and decrease of the cooling capacity of urban vegetation. The simulation results show generally an increase in near-surface moisture when urban development replaced desert land, a decrease in horizontal wind speed and an increase in boundary layer heights when agricultural land was developed.