Dynamic ammonia exchange within a mixed deciduous forest canopy in the Southern Appalachians

Ammonia ( <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> ) concentration and flux measurements made in 2016 in a mixed deciduous forest at the western North Carolina Coweeta Hydrologic Laboratory are analyzed using a multi-layer, one-dimensional column model with detailed canopy physics and bi-directional exchange. Simulations for April 26-30 and July 19-30 are presented to assess the model's ability to represent measured in-canopy <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> profiles and probe the processes that control bi-directional exchange with the canopy and forest floor. During dry canopy conditions, model simulations are found to well reproduce measured in-canopy profiles for both the April and July periods, given appropriate model inputs. Results from the model, and the shape of in-canopy <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> profiles, are sensitive to vertical turbulent mixing, the values of the input soil/litter emission potential, and the assumed litter resistance. <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> fluxes simulated above the forest canopy are very small (-25 to -5 ng m^-2 s^-1 in April and < 1 ng m^-2 s^-1 in July) with primarily deposition to the canopy during the April time period, but with mixed deposition/emission during July. The model also suggests that net deposition or emission of <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> can be a function of location within the canopy, depending on the difference between the air concentration and the effective canopy compensation point. However, during periods when the canopy is wet from overnight dew and drying rapidly, the model does a poor job of replicating in-canopy profiles, typically underestimating <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> concentrations, since the model does not account for the release of <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> from evaporating dew. Although available data during the field campaign are not sufficient to rule out other potential hypotheses, given that the model reasonably reproduces in-canopy profiles during dry canopy periods, but fails during periods of rapid drying, the results are suggestive that dew is playing a major role in <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> concentration changes observed in July during the field study. Additional studies and measurements are needed to determine the processes and environmental controls that affect <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> absorption and release from dew and to evaluate the importance of this process for modeling deposition and re-emission on the regional scale. Further questions that arise from our findings are whether the variation of <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> deposition or emission with location in the canopy is important from an ecological perspective and how in-canopy dynamics might be represented in regional-scale air quality models. Traditional big-leaf approaches of modeling <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> bi-directional exchange cannot account for in-canopy variation such as that presented here, and so multi-layer approaches may need to be developed for more nuanced estimates of <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub><mml:mrow><mml:mi>N</mml:mi> <mml:mi>H</mml:mi></mml:mrow> <mml:mrow><mml:mn>3</mml:mn></mml:mrow> </mml:msub> </mml:math> deposition to forest ecosystems.