The use of inferential models for estimating nitric acid vapor deposition to semi-arid coniferous forests

Urban areas emit significant amounts of pollutants that impact forest ecosystems. One of the most important of these is nitric acid vapor (HNO₃), a nitrogen-containing gas that deposits efficiently to forest canopies. Since measuring HNO₃ fluxes directly is often impractical and costly in remote forest locales, inferential techniques are most often used to estimate HNO₃ flux. Given the highly efficient deposition of HNO₃, many of these inferential models assume that leaf surfaces are a ‘perfect sink’ for HNO₃ (i.e., that resistance to HNO₃ deposition is negligibly small or zero). This study tests the ‘perfect sink’ assumption in an open gas exchange system by exposing Abies magnifica, Abies concolor, and Pinus jeffreyi seedlings to concentrations of 1–13 ppb at 4–20% relative humidity. We find that, at these humidities and concentrations, cuticles are not perfect sinks for HNO₃, with cuticular resistance values ranging from 20 to 184 s m⁻¹. In addition, our results indicate that accumulating HNO₃ on leaf cuticles at these concentrations leads to higher cuticular resistance over 8–12 h exposure periods. Based on this laboratory data, we then parameterized cuticular resistance using a single-layer inferential model for semi-arid forests in the Lake Tahoe Basin. Modeled fluxes using this modification were 33% lower during well-mixed daytime conditions than the fluxes from an identical model run using the perfect sink assumption. Since HNO₃ can often account for more than half of atmospheric deposition, we conclude that inferential models that assume foliage to be perfect HNO₃ sinks are inaccurate, especially in semi-arid forests where significant amounts of HNO₃ can accumulate on leaf surfaces during dry periods.