The efficacy of using CFD generated library databases to describe wind field and dispersion characteristics of major urban areas is examined. These databases might be used to study urban transport and dispersion problems in major cities, or even be applied to plan for, and assess the effects of a release of airborne chemical and/or biological agents. The use of CFD generated library databases could circumvent the usual complaint that detailed CFD modeling of large urban areas is too slow and expensive for such uses. At the same time, these databases would still provide more detailed and accurate information to planners than simpler urban dispersion models.

We will examine the computing and storage requirements for such a database for selected major U.S metropolitan areas, using a CFD code (HIGRAD) that has bee under development at Los Alamos National Laboratory, under the DOE Chemical and Biological Non-Proliferation (CBNP) Program. Such requirements are determined by the minimum number of simulated scenarios that one might use to represent the greatest range of possible - and probable - meteorological conditions. We will also demonstrate this implementation for a major metropolitan area.