An approach to characterize within-grid concentration variability in air quality models

A new methodology to obtain subgrid air quality concentration variability (SGV) as distribution functions is described. Grid-based air quality models provide deterministic single value outcomes and thus do not represent concentration spatial details which can vary widely especially within urban and highly industrialized areas. While such spatial details can be obtained by running grid-based regional-scale air quality models at finer resolution or by employing local-scale models, such implementation may be impractical for performing a variety of applications, e.g., health exposure assessments. Strategically we propose a paradigm that can provide operational AQ models with supplemental estimates of SGV distributions as analytic functions provided on hourly and grid-by-grid bases. We illustrate and discuss this prototype implementation for a case study of benzene for several 12-km grid cells in the Wilmington, Delaware area. For this effort the contribution to SGV is limited to that from fine grid emissions distributions.

► Conceptual framework to provide subgrid details to air quality grid models.
► Paradigm fits two-parameter Weibull distribution to subgrid variance histograms.
► Regression of operational model outputs used to model Weibull parameters for Benzene.
► Methodology provides on-line capability for regional air quality models.