Application of a Coupled Urban Surface Boundary Layer Model to Assess the Impact of Roof Albedo Increase on Low-level Temperature in Chicago

A simple 1-D boundary-layer model (Troen and Mahrt 1986) that treats both local and non-local mixing is coupled with an urban surface scheme (Masson 2000) and a 5-band longwave cooling scheme (Roach and Slingo 1979) in order to simulate the sensitivity of low-level air temperature to proposed modifications of roof albedo in Chicago. This combination of models has been tested against observations from the BUBBLE project and shown to perform well in terms of air temperature and energy balance fluxes. Three-day simulations are conducted for summer, winter and spring conditions for downtown and residential land use areas with varied cloud cover, and roof albedos (ranging from 0.06 to 0.65). Changes in the daily minimum, maximum and average air temperature within the street canyon and above roof level are considered. Results show the largest modeled cooling is a 5°C decrease in maximum daily air temperature due to a 0.06 to 0.65 roof albedo change in the clear summer case in downtown (λ_p = 0.53). At the opposite end of the spectrum, wintertime minimum temperatures with thick cloud cover are essentially unaffected by roof albedo (rooftop snow is ignored). Modelled daily mean clear sky summertime cooling due to the 0.06 to 0.65 roof albedo increase was in the 3.0-3.5°C range for both urban landuses. The summer season daily maximum and average temperature decreases are approximately 50% of their respective single day clear sky decreases. An estimate of the yearly-averaged temperature decrease for this same roof albedo change in the residential scenario (λ_p = 0.33), derived from model output and averaged by the climatological reduction in solar flux by clouds, is approximately 1°C. Changes in yearly HDD and CDD are estimated. Modelled temperature changes are likely to be upper bounding values because the homogenizing influence of advection is ignored.