| 30 Dec 2025
Alkaline dust deposition to foliage surfaces likely enhances the dry deposition velocity of SO2: An investigation in the Alberta Oil-Sands Region using the GEM-MACH air-quality model
Abstract. We examine the potential impact of alkaline particle deposition on foliage and its influence on sulphur dioxide dry deposition velocities, using a new theoretical development, a high resolution air-quality model, and comparisons to observations. Our study domain encompasses the Athabasca Oil Sands Region, an industrial area where base cation-bearing fugitive dust from open-pit mining coexists with elevated SO2 emissions from large stacks. Our pH-modulated dry deposition scheme links thin aqueous films on foliage to local chemical conditions, including alkaline dust accumulation. We present the mechanism's theoretical basis, along with a simplified algorithm predicting foliage water pH and linked to SO2 deposition velocities.
We predict enhanced SO2 deposition, due to increased leaf surface pH from dust co-deposition near major dust sources, often by more than 1 cm s⁻¹. These result in dry deposition fluxes 2.5 to 10 times greater than in the absence of these effects – consistent with estimates from aircraft studies. The enhanced deposition reduces surface SO2 concentrations by up to 60% near sources, improves agreement with continuous monitoring data, and reduces normalized mean bias at several stations. Taylor diagram statistics show improved model temporal variability performance. Further from sources of base-cation-containing dust, aqueous films on foliage remain acidic, reducing SO2 deposition velocity and increasing concentrations.
We make specific recommendations for new observation data which would reduce formulation uncertainties. The findings have broad implications for global SO2 budgets, given the significant role of wind-blown mineral dust in influencing atmospheric acidity and trace gas removal.
Review: Miller et al., 2026
“Alkaline dust deposition to foliage surfaces likely enhances the dry deposition velocity of SO2: An investigation in the Alberta Oil-Sands Region using the GEM-MACH air-quality model”
Summary
This manuscript describes work to improve process representation of SO2 dry deposition in the GEM-MACH regional air quality model. Dry deposition is an important process to capture in chemistry transport models, but process realism is difficult to achieve due the sub-grid scale nature of the contributing processes. The authors address this problem by targeting dry deposition on wet vegetation surfaces. They expand GEM-MACH’s dry deposition mechanism to include a scheme that calculates pH on the thin films of water that accumulate on foliage. They hypothesize that this will improve SO2 dry deposition in regions like the Athabasca Oil Sands, where high atmospheric concentrations of acidic and alkaline gases and aerosols can influence foliage pH, and therefore deposition of acidic and basic gases. The authors find that GEM-MACH can better predict SO2 dry deposition velocities in the Athabasca Oil Sands Region when this scheme is included, although the impacts on atmospheric SO2 concentrations is small.
Overall, the manuscript is well written in clear language. However, I think that the manuscript is over-long relative to the limited scope of the results. It is unclear what the wider impacts, if any, the new foliage water pH scheme would have on e.g. deposition velocities of other acidic and basic gases like NH3 or HNO3 and what impact it might have a dusty or wildfire affected region. While I recommend the manuscript for publication, I have some General and Technical comments below that should be addressed first.
General Comments
Technical Comments
Abstract
Suggest “We examine here the potential for simultaneous…”
Introduction
Section 2.2
Section 3 – Dry deposition algorithms
Here, is “M” mol L-1?
Section 4 - simulating foliage pH
Suggest pointing to the specific section/subsection.
Section 5.1
Section 5.2
Section 5.3
Figures 11 and 12 show that there are only small changes in SO2 concentrations between the base and CALCCO3_Lw_high simulations at most of the sites and only five sites where CALCCO3_Lw_high substantially reduces bias against the observations. This is despite the better agreement between CALCCO3_Lw_high and observed VdSO2 shown in Figure 9. Is there anything linking the five WBEA sites where CALCCO3_Lw_high reduces bias against the observations and Oski-otin, YAJP and DWEF? E.g. location or vegetation type?
I also found Figure 12 hard to interpret due to the small font and symbol sizes. Perhaps the authors could consider i) including just Figure 12b (the results from June that are also presented in Figure 11) and ii) summarizing the results from Figures 12a, c and d?
Can the authors suggest a reason/s why CALCCO3_Lw_high reduces bias against the atmospheric SO2 observations at some sites and months, but not others? Are these sites linked by location or vegetation type? Further, is there anything linking the five WBEA sites where CALCCO3_Lw_high reduces bias against the observations and Oski-otin, YAJP and DWEF? Again, location or vegetation type?