UC Berkeley

An obstacle to the modeling of strategies to mitigate extreme urban temperatures is frequently the lack of on-site meteorological data. The current study thus reports on a method that used the Weather Research and Forecasting (WRF) model to generate inputs for the ENVI-met model to produce building-scale canyon temperatures within a 300 m square near downtown San Jose. A land use distribution was generated for WRF by a WUDAPT classification, and the days of interest were then the hottest day in California history and a typical summer day. The source of meteorological data for ENVI-met, run with a 1.5 m cubic grid, was either an urbanized version of WRF; its default version; or observations at the closest NWS site. All WRF simulations were run on a 1 km grid, and output at its grid closest to the study area provided ENVI-met with lateral boundary conditions. The mitigation strategy was comprised of three parts, which either increased vegetation, rooftop albedo, or architectural shade elements. Results showed all strategies with only negligible impacts on ENVI-met nighttime 1 m level street canyon temperatures. Increased vegetation, however, was the most effective daytime strategy on both days, as it affected the largest area. The maximum vegetative cooling on the extreme and average days was −3.5 and −3.3 °C, respectively. While increased rooftop albedos produced near negligible impacts, increased architectural shading produced corresponding values of −1.6 and −1.7 °C, respectively.