Estimating Sources, Sinks, and Fluxes of Reactive Nitrogen and Sulfur within a Forest Canopy Using Eulerian and Lagrangian Inverse Models

Estimating the source/sink distribution and vertical fluxes of air pollutants within and above forested canopies is critical for understanding the biological, physical, and chemical processes influencing the soil-vegetation-atmosphere exchange. The vertical source-sink profiles of reactive nitrogen and sulfur were examined using multiple inverse modeling methods in a mixed hardwood forest in the southern Appalachian Mountains, which is a region sensitive to acidic deposition. Measurements of the vertical concentration profiles of ammonia (NH₃), nitric acid (HNO₃), sulfur dioxide (SO₂), and ammonium (NH₄⁺), nitrate (NO₃^-), and sulfate (SO₄^2-) in PM_2.5 were measured at the Coweeta Hydrologic Laboratory site during five study periods between May 2015 and August 2016. The mean concentration of NH₃ decreased with height in the upper canopy and increased below the understory toward the forest floor, indicating that the canopy was a sink for NH₃ but the forest floor was a source. All other species exhibited patterns of monotonically decreasing concentration from above the canopy to the forest floor.

Using the measured concentration profiles and within-canopy flow fields, we estimated the vertical source-sink flux profiles using three inverse approaches: a Eulerian high-order closure model (EUL), a Lagrangian localized near-field (LNF) model, and a new full Lagrangian stochastic model (LSM). The models were run to estimate sensible heat flux profiles from multilevel temperature measurements, which were evaluated using within- and above-canopy eddy covariance sensible heat flux measurements. Despite fundamental differences in the modeling framework and assumptions, all three models reproduced the shape and the magnitude of the measured sensible heat flux profile reasonably well, which is an important step in developing confidence in calculated flux profiles of reactive compounds.

Models predicted positive (upward) NH₃ fluxes near the forest floor, indicating emissions from litter or soil. The modeled above-canopy flux of NH₃ tended to be negative (downward), indicating that the forest was a net sink of NH₃. The modeled flux profiles of HNO₃ and SO₂ presented a monotonically decreasing trend with most uptake occurred in the upper canopy. Significant differences in the estimated flux profiles can be found between different models and for different timescale inputs. This presentation examines aspects of model validation against measured sensible heat fluxes, model sensitivities to turbulence time scale and other model inputs, and features of the flux profiles predicted for reactive compounds. The knowledge gained in this study will benefit the development of soil-vegetation-atmosphere models capable of partitioning canopy-scale deposition of nitrogen and sulfur to specific ecosystem compartments.