The development and release of the AERSURFACE landuse preprocessing program forAERMET/AERMOD has substantially changed the process for determining the landusecharacteristics (i.e. surface roughness, albedo, and Bowen ratio) that surround ameteorological monitoring site. Previous modeling guidance recommended that a 3-kilometer (km) upwind fetch in wind-direction-based sectors be considered in order todetermine appropriate values for the surface roughness, albedo and Bowen Ratio. TheAERMOD Implementation Guide and AERSURFACE User’s Guide now propose thatwith respect to surface roughness, a smaller upwind fetch (~ 1km) more appropriatelyrepresents the turbulent processes that may impact conditions at the meteorologicalmonitor. The use of this smaller domain, along with the use of the digital National LandCover Database (NLCD, version 1992), has resulted in much different values for sectorspecificsurface roughness than those derived from the 3-km methodology.
From an air quality permitting perspective, these significant changes in determiningsurface roughness have not only made it difficult for modelers to show clearrepresentativeness between facility landuse and meteorological station landuse, but it hasalso caused AERMOD modeled concentrations to change significantly in many cases.For example, this AERMOD model sensitivity to landuse variability can result in projectsmodeling above the modeling significance levels (MSL), triggering NAAQS and PSDIncrement modeling requirements, whereas previous impacts were shown to beinsignificant.
This paper presents a comparison of model predictions from the same hypotheticalfacility, given differing landuse characterizations. While surface roughness variabilityappears to result in the most model sensitivity, model sensitivity to Bowen ratio is alsoconsidered since many state and local agencies have proposed differing approaches forquantifying soil moisture availability, which has a potentially significant effect on valuesassigned for that parameter.