Improved representation of the urban radiative budget in mesoscale atmospheric models

Mesoscale atmospheric models are commonly used in urban meteorology applications including air quality and heat island studies. These models typically have been developed to be applied at scales too large to effectively resolve urban features. The available surface data as well as limitations in the parameterizations used to represent heat, mass, and momentum exchange at the surface necessitate large grid cells. As a result, the urban landscape is commonly represented by individual grid cells on the order of 2 to 5 km on a side. Within an individual grid cell the urban surface is modeled as a flat surface at a particular altitude. Furthermore, the radiative properties of the grid cell are estimated as the average of the surfaces making up the grid cell. While this process is limited by data availability and quality it also suffers from the assumption that the incident radiation is from directly above the surface. In actuality as the sun traverses the sky the effective albedo of an urban grid cell can change dramatically. In the morning and evening the effective urban albedo is dominated by the albedo of exterior building walls. Even at mid-day, the role of street-level surfaces is much less important than might be estimated by their fractional area coverage. Furthermore, reradiation within the urban canopy can significantly reduce the overall effective albedo. This paper will present an analysis of how the effective albedo of typical urban grid cells can vary diurnally and annually as a result of these effects. Results will be presented in terms of the effect on the integrated energy balance of the city as well as the implications for modeling studies that require accurate representation of the urban energy balance.