The Predictive Value of the Ventilation Index for Atmospheric Dispersion in Complex Terrain

The predictive value of the Ventilation Index (VI, also known as Clearing Index) for atmospheric dispersion in complex terrain is assessed using a Lagrangian particle dispersion model –FLEXPART, coupled with a mesoscale weather prediction model – the Advanced Regional Prediction System (ARPS). Calculated as the product of the mixed layer depth and transport wind speed, VI is used to indicate the potential for the atmosphere to disperse airborne pollutants from a stationary source. As a standard product of air quality and fire weather forecasts, VI has been used by air resource and fire managers to help determine when air quality and visibility might be affected by smoke from wildland fires. Despite its popularity, there have been few assessments of the predictive value of VI in different regions, especially regions with complex terrain.

In this study, how well the VI represents the ability of the atmosphere to disperse pollutants is assessed by tracking the movement of particles as they are released from different areas in complex terrain. Numerical simulations using FLEXPART driven by ARPS are carried out in the Lehigh Gap region of eastern Pennsylvania. The topography of the region is dominated by the Blue Mountain, a west-southwest – east-northeast oriented ridge with a gap in the middle where the Lehigh River flows through. Meteorological fields simulated by ARPS are used to drive FLEXPART simulations of particle transport and dispersion. More than 250,000 particles with properties representative of particulate matter of diameter 2.5 mm or smaller (PM2.5) are released over a 7-h-period across the model domain. Five one-square-kilometer boxes, positioned in areas of the domain with different terrain characteristics, are utilized to examine how the VI performs in and around areas of complex terrain. For each box, the ensemble-mean residence time of all particles released in that box is compared to the VI values averaged across the box. Results suggest that while the atmospheric transport and dispersion regime indicated by the calculated VI values are consistent with the behaviors of particles released on the windward side and ridge top and in areas away from the mountain, VI poorly represents the actual transport and dispersion pattern of particles released on the leeward side of the mountain, where localized flows play an important role in modulating particle dispersion.