Resumen
In this study, we examined how modeled ozone concentrations respond to changes in anthropogenic emissions when different modeled emissions of biogenic volatile organic compounds (BVOCs) are used. With biogenic emissions estimated by the Model of Emissions of Gases and Aerosols from Nature (MEGAN) and the Biogenic Emissions Inventory System Version 3 (BEIS3), the Community Multi-scale Air Quality with the High-order Direct Decouple Method (CMAQ-HDDM) simulations were conducted to acquire sensitivity coefficients. For the case study, we chose 17?26 August 2007, when the Southern Korean peninsula experienced region-wide ozone standard exceedances. The results show that modeled local sensitivities of ozone to anthropogenic emissions in certain NOx-saturated places can differ significantly depending on the method used to estimate BVOC emission, with an opposite trend of ozone changes alongside NOx reductions often shown in model runs using MEGAN and BEIS3. Findings of increased ozone concentrations with one model and decreased ozone concentrations with the other model implies that estimating BVOCs emissions is an important element in predicting variability in ozone concentration and determining the responses of ozone concentrations to emission changes, a discovery that may lead to different policy decisions related to air quality improvement. Quantitatively, areas in the 3-km modeling domain that experienced daily maximum one-hour ozone concentrations greater than 120 ppb (MDA1O3) showed differences of over 20 ppb in MDA1O3 values between model runs with MEGAN and BEIS3. For selected monitoring sites, the maximum difference in relative daily maximum eight-hour ozone concentrations (MDA8O3) response between the methods to model BVOCs was 4.2 ppb in MDA8O3 when we adopted a method similar to the Relative Reduction Factor used by the US Environmental Protection Agency (EPA).