| 22 Oct 2025
Modification and validation of a commercial dynamic chamber for reactive nitrogen and greenhouse gas flux measurements
Abstract. Reactive nitrogen compounds (NO, NO2, HONO, NH3 and others; Nr) play important roles in atmospheric processes, and their cascading impacts throughout the Earth system have adverse effects on both the environment and human health. The fluxes of these gases at the surface-atmosphere interface have been studied in isolation or in smaller subsets by micrometeorological techniques or chambers, but simultaneous observations of all Nr species alongside standard greenhouse gases (GHGs) as a function of time have not been reported. Here, a dual-dynamic chamber system was developed for Nr by modifying a commercially available system for GHG fluxes for use with destructive analyzers and to account for chemical changes. The resulting platform makes the measurement of Nr and, by extension other reactive gases, more widely accessible to the scientific community, as custom chambers do not need to be fabricated.
System modifications to passivate surfaces were implemented, so that Nr gases like NO2 could be effectively transferred to standard gas analyzers, with an initial 36 % loss due to transformations ultimately minimized below analyzer detection limits (~10 %) under relevant atmospheric conditions. The modified 72 L chamber did not see a change in the baseline response times for GHGs or NO at a flow rate of 2 L min-1. They retained the same values as an ideal non-reactive trace gas (τ = 37–39 min versus 36 min. The modifications improved the transfer time constants of NO2, HONO, and NH3 by up to 2 min, but substantial surface interactions for NH3 remain. In all cases, a surface interaction term needs to be characterized for these gases to obtain accurate fluxes. Losses of NO2 and O3 by known gas phase reactions, or from deposition and reaction on pristine and aged chamber surfaces, were characterized across a range of environmentally relevant relative humidities (RH) and mixing ratios. The final dual-chamber system configuration includes a measurement and reference chamber, which are necessary to implement the corrections for surface effects and chemical transformations when accurately quantifying dynamic fluxes via a mass balance framework.
Proof-of-concept measurements of Nr fluxes from agricultural soil samples under controlled lab conditions as a function of soil water content were able to quantify emissions of NO, NO2, HONO, NH3, and N2O simultaneously, when subject to fertilization experiments using urea, ammonium carbonate and bicarbonate, and ammonium nitrate. Unfertilized replicate agricultural soil samples showed variability in NO2 and HONO emissions when prepared with minimal disturbance to the soil structure, with values consistent with those reported by in-situ field measurements. These oppose maximum potential fluxes characterized in prior lab soil manipulations, particularly for HONO relative to NO. Last, fluxes were quantified with destructive gas analyzers in the field with the dual-chamber system on an in-use agricultural soil and included a urea-based fertilizer perturbation to stimulate microbial and chemical transformation and transfer Nr to the atmosphere. The resulting fluxes observed show good agreement with prior reports based on other flux techniques. The mass balance terms within the dual-chamber approach are fully inspected from the pilot deployment in the field, along with an error analysis, to aid in the uptake of this approach by the community.
Competing interests: TCV received supporting in-kind funds for this work from Eosense, Inc. and Picarro as it is mandatory in the NSERC Alliance Missions programme funding structure which facilitates research partnerships between the academy and industry. NN, CC, and SE are employed by Eosense, Inc.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.
Summary:
This manuscript details the experimental design, configuration/modifications, and utility of a dual system of commercial dynamic chambers for simultaneously quantifying emissions fluxes of reactive nitrogen species and greenhouse gases. This work carefully outlines the system details as well as the many considerations that go into the flux calculations, which are further described in the context of a proof-of-concept experiment aimed at quantifying emissions fluxes of nitrogen species and greenhouse gases from agricultural soils. This manuscript is a highly valuable addition to the literature because it provides: 1) a description of the dual-chamber system detailed enough for others to reproduce, 2) a comprehensive account of the considerations and calculations needed for dual-dynamic chamber flux calculations, and 3) initial proof-of-concept results from controlled laboratory and field experiments. The latter shows great promise in being able to set a foundation for charactering changes in emissions fluxes of nitrogenous species and greenhouse gases from agricultural soils as a function of relative humidity, temperature, soil moisture, fertilizer applications, and more, which are critical for understanding emissions from food systems that impact air pollution and climate. This manuscript is well written, and it fits well into AMT. I recommend this manuscript for publication. Please find a few minor comments and suggestions below.
General comments:
While this paper is very comprehensive, a missing component of this work seems to be a brief, clear, and upfront discussion of the advantages of this dual-dynamic chamber system compared to the system described in Pape et al., 2009, who also implemented a blank chamber in reference to up to as many as 6 sampling chambers and measured Nr species. It would strengthen this work to be more explicit about the novel aspects of this dual-dynamic chamber system and proof-of-concept experiments in comparison to the dynamic chamber system and field measurements described in Pape. This could be integrated into the last 1-2 paragraphs of the introduction.
At first look, the 30-minute sampling cycle for the MC and RC chambers seems a bit short. On the other hand, a longer sampling time probably allows for more dilution, reactivity, etc. It seems like there is a delicate balance of the time to observe enhancements in the chamber, trace gas analyzer response times, chamber flush times/flow rates, dilution times from the replacement make-up gas, and the time associated with wall loss in the chamber, etc. Is it possible to combine all of these timescales into a single schematic to demonstrate this balance? Is there a sweet spot in the schematic that clearly stands out for this set up? At a minimum it might help to add some more detailed explanation about why the chamber time to reach a reasonable detection limit for HONO (e.g., Section 2, L250) is used compared to the longer chamber fill/empty time for sticky NH3 (e.g., Figure 2 and Table 1). How would using a different chamber sampling time impact your flux results? It seems like a good bit of the chamber timing characterization was also discussed in Pape et al., 2009 (e.g., their figures 6 and 7). Briefly summarizing the timing simulations in Pape with respect to your timing scheme could help clarify and strengthen this work.
This paper is well-written and comprehensive. However, it is a bit cumbersome in places. A suggestion for improved readability could be to reconsider what goes in the main paper vs. the SI. For example, Section 2.1.2 seems better suited for the SI, while important analyzer calibrations and gas handling considerations are handy to have in the main paper. Similarly, important equations and calculations, such as those in Section S7, while cumbersome for the main paper, could go into an appendix of the main paper for increased visibility.
Specific comments:
L183: Suggest changing “manipulated” to “performing controlled experiments”
L250: Is there a reason that HONO is used to determine the sampling time interval, compared to say NH3? Section 2.6 did not seem to provide further explanation for this selection. Is this somewhere else? At a minimum, this section seems like a good place to summarize some of the chamber timing simulations in Pape et al., 2009. It would also be a good place to discuss similarities and differences in timing between your experiments and their studies (e.g., differences in timing with respect to differences in flow rates).
L279: This should be Section 2.1.3, right?
L281: Section 2.1.2 seems like it could go in the SI, and then the instrument descriptions can naturally follow on Section 2.1.1.
L305-317: Can you determine an analyzer detection limit for HONO?
Section 2.1.3 and L309: Did you actually measure interferences from other oxidized nitrogen species in the NO2 measurement?
L329: The Picarro analyzer response times are given as 8 s for non-sticky species, which are assumed to be 1/e times of a single exponential function. While this seems sufficient compared to the chamber empty/fill times for non-stick species, the longer analyzer response time for sticky NH3 (reported as <2 min) may rival the chamber fill/empty times (e.g., Table 1) when a double exponential fit and a longer, more appropriate, signal recovery time (e.g., t90, the time for 90% signal recovery) for sticky molecules are considered. Can you comment on this?
While the chamber response times are well-addressed in Section S3, it was not obvious how the analyzer’s response times are characterized for stick molecules (NH3 and HONO). What is the modified NOx analyzer’s response time for the HONO measurement and how is it defined?
Another consideration is whether the time responses of these analyzers and the chambers were characterized multiple times throughout the study period to ensure that they did not change. The time response of instrument systems measuring sticky molecules are well known to degrade over time and usage as sampling surfaces get dirty or saturated with moisture (e.g., Ellis et al., 2010; Moravek et al., 2019; Roscioli et al., 2016, https://pubs.acs.org/doi/10.1021/acs.jpca.5b04395, etc.). How might changes throughout the study period in analyzer and chamber response times for sticky molecules impact your results? Some additional discussion could help clarify. (See also additional related comments below with respect to lambda.)
L337: I agree that full span checks for greenhouse gases are not regularly needed for the Picarro, but I’m not sure that holds true for sticky NH3. You mention performing regular calibration checks anyway, which is good. How often were those performed? Did you observe changes, or more variability, in the NH3 channel? How about checks of the instrument time response, especially to NH3? Did that time response change over the duration of your study?
Related details about analyzer calibrations (in the SI) could also be clarified. Specifically, how was the modified NOx analyzer calibrated for HONO and how was the Picarro analyzer calibrated for NH3. This should be added to Section S2 if different from the methods used to characterize the chamber time response measurements in Section S3. If they are the same, it might be clearer to add the calibration procedure to section S2 and then refer to it in section S3.
L354 and Section 2.2: It would help to see a photo of the chambers in the lab or in the field in the main paper in this section. Maybe a photo of the chamber set up in the lab or field (like in Figure S2 - very cool!) could be added to Figure 1 as a part C in the main paper?
The PFA film over the lid is a smart addition to reduce surface interactions. Is the air sample completely isolated from the polyacrylate cover (aka. is there a leak-free seal between the PFA film and the cover)? If not leak-free, does the adhesive on the double-sided sticky tape create any chemical interference? How are the chamber fill/empty times impacted by sample air getting between the PFA film and the lid? Could these factors impact lambda?
Section 3.1.2: It would help to add callouts in this section, like where D and NO2 reactivity are discussed, to their respective details in the SI (S3 and S7).
L630: Can you determine an uncertainty for Lambda?
Table 1 shows substantial differences in %D for the chamber empty vs. chamber fill times for NH3. Does %D change over the duration of the study period as the chamber gets dirty with usage? By the same reasoning, can or does lambda change over the duration of the study period (e.g., can or does Figure S11 change over time throughout the study period as the chambers are used and get dirty)?
Figure 4: This is a nice figure. It’s a bummer that there are no N2O or NH3 measurements for the samples with “none” and “AN” fertilizer applications, as N2O and NH3 fluxes are predominant from the other treatments. Will a comparison of soils and treatments using this system be included in your forthcoming paper about the field study? That would also be an exciting addition to the literature. :)
Figure 5: The timescales in this figure could use some clarification, and this might be easily solved by simply adding some labels to each panel. These labels could simply say “after fertilization” for the NH3 time series and “before fertilization” for the NO and N2O timeseries. It is also unclear why the dates and times are not the same (e.g., 2 pm versus 5 am on 9/5/2022) for simultaneously measurements of NO and N2O during the “before fertilization” sampling. Is this because of differences in peak emission times as shown in Figure 3?
Section 3.3.3: How would propagating an uncertainty in lambda impact the total uncertainties in the fluxes, especially for NH3? Since lambda is time dependent (e.g., Figure S11), can you comment on how a longer chamber sampling interval might impact the fluxes, again especially for NH3.
Sections S2 and S3: It seems that the instrument calibration methods for NH3 and HONO are missing from Section S2. It could be the case that the HONO and NH3 calibration sources used in Section S3 are also used to calibrate the analyzers and characterize instrument time responses. If that is the case, then I suggest reorganizing these sections to clarify. Along the same lines, section S2 should also describe how the analyzer response times are characterized. (Sorry if this is a repeat comment from above.)
Section S7: It is helpful for readability if you can redefine parameters in the equations (like A, Q, F, c, R, X, t1, and t2, etc.) for the equations in the SI.
Figure S9: Suggest adding the absorption cross section spectrum for NO2 to this figure for comparison. Can you estimate how modifying the chamber with the PFA film attached to the lid might impact loss by photolysis? Is there a transmission curve somewhere for this commercial material?