the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Assessing representativity of NH3 measurements influenced by boundary-layer dynamics and turbulent dispersion of a nearby emission source
- 1Wageninen University & Research, P.O. Box 47, 6700, AA Wageningen, the Netherlands
- 2National Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721, MA Bilthoven, the Netherlands
- 1Wageninen University & Research, P.O. Box 47, 6700, AA Wageningen, the Netherlands
- 2National Institute for Public Health and the Environment (RIVM), Antonie van Leeuwenhoeklaan 9, 3721, MA Bilthoven, the Netherlands
Abstract. This study presents a fine scale simulation approach to assess the representativity of ammonia (NH3) measurements in proximity of an emission source. Close proximity to emission sources (< 5 km) can introduce a bias in regionally representative measurements of the NH3 molar fraction and flux. Measurement sites should therefore be located a significant distance from emission sources, but such requirements are poorly defined and can be difficult to meet in densely agricultural regions. This study presents a consistent criterium to assess the regional representativity of NH3 measurements in proximity of an emission source, calculating variables that quantify the NH3 plume dispersion using a series of numerical experiments at a fine resolution (20 m). Our fine scale simulation framework with explicitly resolved turbulence enables us to distinguish between the background NH3 and the emission plume, including realistic representations of NH3 deposition and chemical gas-aerosol transformations. We introduce the concept of blending-distance, based on the calculation of turbulent fluctuations, to systematically analyze the impact of the emission plume on simulated measurements, relative to this background NH3. This sensitivity analysis includes systematic experiments varying meteorological factors, emission/deposition and NH3 dependences. Considering these sensitivities, we find that NH3 measurements should be located at a minimum distance of 0.5–2.5 km and 1–3.5 km from an emission source, for NH3 molar fraction and flux measurements respectively. The simulation framework presented here can easily be adapted to local conditions and paves the way for future ammonia research at high spatio-temporal resolution.
Ruben B. Schulte et al.
Status: closed
-
RC1: 'Comment on acp-2021-907', Anonymous Referee #1, 17 Dec 2021
The manuscript "Assessing representativity of NH3 measurements influenced by boundary-layer dynamics and turbulent dispersion of a nearby emission source" feeds in an actual scientific debate on new ammonia emission methods which encompass precision and sensitivity. Several key drivers were consider in the simulation work and the introduction of new variables such as blending-distance represent a valuable contribution to the study.
A general shortcoming of this submission is, that a short period of time was considered in a very specific context were orography plays no role in the analysis. This study limitation is mentioned in discussion section but it would be interesting discuss a little bit further the representativeness of the conclusions achieved in a different orographic context, the variability throughout 24 hours due to atmospheric estability changes and the potential effect of this variables in blending-distance variability. This study is limited to 3 hours with 30 minute NH3 flux input (6 values) during the central hours of the day which seems a little bit restricted. It would be interesting for future works to extend this analysis period.
It is kind of mention in the study but it results a little bit confusing the mixture between local emission sources and regional emission in a densely agricultural area. Distances over 1 km are not easily achievable without any ammonia emission sources in this areas. The study uses farms as emission sources, but in this areas crop fields are also diffusion emission sources that may deeply influence the monitoring grid design and conclusions achieved in relation to minimum distances required in this work and alter plume behaviour significantly specially in fertilizing season. It would be nice to see any impression regarding point sources and non-point sources of emission when applying this model.
It would be useful a figure showing the different spatial configuration (emission sources and measuring sites) tested in the simulation with the model, as well as information about the weather conditions (temperature, wind speed, wind direction, etc.) at which the simulations were performed during the experimental period (15:00 to 17:00 hours).
I find especially interesting the analysis of the effect of the impact of the emitted plume in the average ammonia values testing different values. The analysys of the effect of background NH3 concentration in the blending-distance it provides valuable information from a regional perspective. Further discussion on this may be interesting.
The study state the relevance of the distance needed from the measurement sites to the emission source in order to avoid bias but it would be advisable to mention how the orography of the area and changes in wind direction (not only wind speed) may affect to this parameter. It would be also interesting to establish the any minimum requirements that should be accomplish or delimit the specific physical context of the site in which the starting hypothesis and study conclusions are valid. If the models do not take into account this aspect it should be mentioned. The replicable capacity of the model and the validity of the minimum distance recommendations concluded may requires much more consideration in the discussion of the results.
Results and discussion sections are a little bit mixed. The result section contains discussion that may be reallocated in section 4 (i.e lines from 324 to 349).
Typing errors:
L43. Agricultuural --> agricultural
L144 intermittenct --> intermittence
L429. Virutal --> Virtual
-
AC1: 'Reply on RC1', Ruben Schulte, 23 Dec 2021
We thank Referee #1 for their thorough reading and comments to our manuscript. The Referee raises several interesting points about the limitations and applicability of the results presented in the study. Below, we would like to give an early response to several of the comments raised by the Referee. We will respond to all points in more detail with the revised manuscript.
- The Referee raises a very interesting point about the representativeness of the results in a different orographic context. This was not considered in the manuscript due to our focus on the Netherlands and the limitations of the model, which requires a flat surface. However, it is possible to represent orographic processes in the model with advection and/or a tilted surface. We are currently looking into how we can best address the points raised by the Referee.
- The Referee suggests to extend the analysis period over which the blending-distance is calculated. We agree that this would be interesting and we intend to perform an additional experiment where the emissions are released at the start of the simulation (8:00 CEST), to analyze the results between 8:00 and 17:00 CEST. Such an experiment could give valuable context to the applicability of the results presented in the manuscript.
- The Referee suggests to show information on different wind directions and different spatial configurations of the measurement sites. We plan to add a new figure where we show the blending-distance for different wind directions. Here, different wind directions are represented by the position of the grid cells with respect to the plume centreline, as each grid cell represents a potential measurement site in the simulation framework.
- Finally, the following comment by the Referee was not fully understood: “It would be also interesting to establish the any minimum requirements that should be accomplish or delimit the specific physical context of the site in which the starting hypothesis and study conclusions are valid. If the models do not take into account this aspect it should be mentioned.” We kindly ask if the Referee could clarify this comment.
-
AC1: 'Reply on RC1', Ruben Schulte, 23 Dec 2021
-
RC2: 'Comment on acp-2021-907', Anonymous Referee #2, 16 Feb 2022
This a well written manuscript with a clear scope.
My main comment regards the basic assumptions in the manuscript. There are two fundamental assumptions in the manuscript: 1) the NH3 concentration in the plume is treated as a decaying scalar decoupled from the background and characterized with a “typical” percentage conversion rate, 2) the same conversion rate is applied for the plume and the background. However, the near source volume interested by the plume dispersion may have a rather different (non-linear) reaction environment and the conversion rate may be locally different. Given the simulation context, the approximations above are needed, and a full check of their validity would perhaps require the simulation of NH3 as a reactive scalar with an appropriate reaction mechanism. Nonetheless, given the experience of the authors in the subject of atmospheric turbulence chemistry interaction, it would be nice to add in the manuscript some more discussion about these assumptions.
Other comments
1) The scenarios simulated by the authors are analyzed separately. I think it would be nice to have some representation (or at least discussion) of the variability implied by the different scenarios if they are combined, e.g., large emission rate, geostrophic wind and low background simultaneously.
2) Line 227. I think it is misleading to refer to turbulent fluctuations as noise in an observation. I suggest removing the sentence.
3) Line 235. The authors calculate fI only for values of the mean concertation above a fixed threshold. This is ok but it would be useful to write down how this threshold is significant compared to the local maximum plume mean concentration at the various downwind positions.
4) Line 379. Although I think that I understand how the authors give the estimate 6-15km, it would be useful a more detailed explanation for the less acquainted readers.
5) Line 420-423 are a repetition of the lines 415-418, please remove it.
-
AC2: 'Reply on RC2', Ruben Schulte, 24 Feb 2022
We thank Referee #2 for their thorough reading and comments to our manuscript. The Referee raises a good point on the two assumptions regarding chemical conversion of ammonia and provides some interesting suggestions to improve the manuscript. Below, we would like to give an early response to several of the comments raised by the Referee. We will respond to all points raised in more detail alongside the revised manuscript.
- The Referee suggests to discuss the implications of the assumptions of a “typical” percentage chemical reaction rate, applied equally to the in-plume and background ammonia. Following the advice of the Referee, we will discuss the implications of these assumptions in the revised manuscript. Here, we will discuss the role of macromixing (plume meandering) and micromixing (in-plume mixing) in the dispersion of the emitted ammonia and its relation to the chemical reaction rate, following Vila et al. (1990).
- We agree with Referee #2 that it would be nice to have some representation of combined scenarios in addition to the sensitivity study presented in Section 3.3, where we identify the driving variables of the blending-distance. We aim to address this comment in the revised manuscript, either in a new subsection or as an Appendix. We are considering performing one or more new experiments where we vary a combination of at least 2 of the variables discussed in section 3.3.
- The Referee comments that it is misleading to refer to turbulent fluctuations as noise in an observations and we fully agree. That was one of the messages in Section 3.1, but it was not properly worded. We can imagine that the magnitude of these turbulent fluctuations and the impact entrainment can have close to the surface might come as a surprise for the ammonia measurement community. One of the advantages of an LES is the ability to visualize this turbulent mixing. In Section 3.1 (Fig. 2) we wanted to show that fluctuations as a result of turbulent mixing can have a large amplitude (>4 ppb) and be quite long lasting (up to 5 minutes), even in the background ammonia concentration. We will take a close look at the wording of this section and are considering removing the sentence mentioning instrumental noise to make sure our message cannot be misinterpreted.
- We are not sure if we correctly understand the comment of the Referee about Line 235. The fluctuation intensity (fI) is calculated for all concentrations, but we subtract a moving average concentration to remove the trend from the ammonia concentration. Next, we compare the fI of the total NH3 (background + plume) to the fI of the background NH3 by calculating the difference in percentages. The blending-distance is calculated by calculating the maximum distance at which the percentage changein fI (PCfI in Fig. 4) reaches a fixed threshold, e.g. a 25 % increase in fI resulting from the emission plume.
We interpret this comment as Referee #2 suggesting that it would be interesting to see how the change in fI and the blending-distance relate to the mean concentration of the emission plume. In other words, how much of an increase in the NH3 concentration is found close to the emission source. We are considering to address this question with a new figure showing the concentration of the emitted NH3 (NH3, plume).Reference:
Vila-Guerau de Arellano, J., Talmon, A.M., and Builtjes, P. J.: A chemically reactive plume model for the NO-NO2-O3 system, Atmospheric Environment. Part A. General Topics, 24, 2237–2246, https://doi.org/10.1016/0960-1686(90)90255-L, 1990.
-
AC2: 'Reply on RC2', Ruben Schulte, 24 Feb 2022
Status: closed
-
RC1: 'Comment on acp-2021-907', Anonymous Referee #1, 17 Dec 2021
The manuscript "Assessing representativity of NH3 measurements influenced by boundary-layer dynamics and turbulent dispersion of a nearby emission source" feeds in an actual scientific debate on new ammonia emission methods which encompass precision and sensitivity. Several key drivers were consider in the simulation work and the introduction of new variables such as blending-distance represent a valuable contribution to the study.
A general shortcoming of this submission is, that a short period of time was considered in a very specific context were orography plays no role in the analysis. This study limitation is mentioned in discussion section but it would be interesting discuss a little bit further the representativeness of the conclusions achieved in a different orographic context, the variability throughout 24 hours due to atmospheric estability changes and the potential effect of this variables in blending-distance variability. This study is limited to 3 hours with 30 minute NH3 flux input (6 values) during the central hours of the day which seems a little bit restricted. It would be interesting for future works to extend this analysis period.
It is kind of mention in the study but it results a little bit confusing the mixture between local emission sources and regional emission in a densely agricultural area. Distances over 1 km are not easily achievable without any ammonia emission sources in this areas. The study uses farms as emission sources, but in this areas crop fields are also diffusion emission sources that may deeply influence the monitoring grid design and conclusions achieved in relation to minimum distances required in this work and alter plume behaviour significantly specially in fertilizing season. It would be nice to see any impression regarding point sources and non-point sources of emission when applying this model.
It would be useful a figure showing the different spatial configuration (emission sources and measuring sites) tested in the simulation with the model, as well as information about the weather conditions (temperature, wind speed, wind direction, etc.) at which the simulations were performed during the experimental period (15:00 to 17:00 hours).
I find especially interesting the analysis of the effect of the impact of the emitted plume in the average ammonia values testing different values. The analysys of the effect of background NH3 concentration in the blending-distance it provides valuable information from a regional perspective. Further discussion on this may be interesting.
The study state the relevance of the distance needed from the measurement sites to the emission source in order to avoid bias but it would be advisable to mention how the orography of the area and changes in wind direction (not only wind speed) may affect to this parameter. It would be also interesting to establish the any minimum requirements that should be accomplish or delimit the specific physical context of the site in which the starting hypothesis and study conclusions are valid. If the models do not take into account this aspect it should be mentioned. The replicable capacity of the model and the validity of the minimum distance recommendations concluded may requires much more consideration in the discussion of the results.
Results and discussion sections are a little bit mixed. The result section contains discussion that may be reallocated in section 4 (i.e lines from 324 to 349).
Typing errors:
L43. Agricultuural --> agricultural
L144 intermittenct --> intermittence
L429. Virutal --> Virtual
-
AC1: 'Reply on RC1', Ruben Schulte, 23 Dec 2021
We thank Referee #1 for their thorough reading and comments to our manuscript. The Referee raises several interesting points about the limitations and applicability of the results presented in the study. Below, we would like to give an early response to several of the comments raised by the Referee. We will respond to all points in more detail with the revised manuscript.
- The Referee raises a very interesting point about the representativeness of the results in a different orographic context. This was not considered in the manuscript due to our focus on the Netherlands and the limitations of the model, which requires a flat surface. However, it is possible to represent orographic processes in the model with advection and/or a tilted surface. We are currently looking into how we can best address the points raised by the Referee.
- The Referee suggests to extend the analysis period over which the blending-distance is calculated. We agree that this would be interesting and we intend to perform an additional experiment where the emissions are released at the start of the simulation (8:00 CEST), to analyze the results between 8:00 and 17:00 CEST. Such an experiment could give valuable context to the applicability of the results presented in the manuscript.
- The Referee suggests to show information on different wind directions and different spatial configurations of the measurement sites. We plan to add a new figure where we show the blending-distance for different wind directions. Here, different wind directions are represented by the position of the grid cells with respect to the plume centreline, as each grid cell represents a potential measurement site in the simulation framework.
- Finally, the following comment by the Referee was not fully understood: “It would be also interesting to establish the any minimum requirements that should be accomplish or delimit the specific physical context of the site in which the starting hypothesis and study conclusions are valid. If the models do not take into account this aspect it should be mentioned.” We kindly ask if the Referee could clarify this comment.
-
AC1: 'Reply on RC1', Ruben Schulte, 23 Dec 2021
-
RC2: 'Comment on acp-2021-907', Anonymous Referee #2, 16 Feb 2022
This a well written manuscript with a clear scope.
My main comment regards the basic assumptions in the manuscript. There are two fundamental assumptions in the manuscript: 1) the NH3 concentration in the plume is treated as a decaying scalar decoupled from the background and characterized with a “typical” percentage conversion rate, 2) the same conversion rate is applied for the plume and the background. However, the near source volume interested by the plume dispersion may have a rather different (non-linear) reaction environment and the conversion rate may be locally different. Given the simulation context, the approximations above are needed, and a full check of their validity would perhaps require the simulation of NH3 as a reactive scalar with an appropriate reaction mechanism. Nonetheless, given the experience of the authors in the subject of atmospheric turbulence chemistry interaction, it would be nice to add in the manuscript some more discussion about these assumptions.
Other comments
1) The scenarios simulated by the authors are analyzed separately. I think it would be nice to have some representation (or at least discussion) of the variability implied by the different scenarios if they are combined, e.g., large emission rate, geostrophic wind and low background simultaneously.
2) Line 227. I think it is misleading to refer to turbulent fluctuations as noise in an observation. I suggest removing the sentence.
3) Line 235. The authors calculate fI only for values of the mean concertation above a fixed threshold. This is ok but it would be useful to write down how this threshold is significant compared to the local maximum plume mean concentration at the various downwind positions.
4) Line 379. Although I think that I understand how the authors give the estimate 6-15km, it would be useful a more detailed explanation for the less acquainted readers.
5) Line 420-423 are a repetition of the lines 415-418, please remove it.
-
AC2: 'Reply on RC2', Ruben Schulte, 24 Feb 2022
We thank Referee #2 for their thorough reading and comments to our manuscript. The Referee raises a good point on the two assumptions regarding chemical conversion of ammonia and provides some interesting suggestions to improve the manuscript. Below, we would like to give an early response to several of the comments raised by the Referee. We will respond to all points raised in more detail alongside the revised manuscript.
- The Referee suggests to discuss the implications of the assumptions of a “typical” percentage chemical reaction rate, applied equally to the in-plume and background ammonia. Following the advice of the Referee, we will discuss the implications of these assumptions in the revised manuscript. Here, we will discuss the role of macromixing (plume meandering) and micromixing (in-plume mixing) in the dispersion of the emitted ammonia and its relation to the chemical reaction rate, following Vila et al. (1990).
- We agree with Referee #2 that it would be nice to have some representation of combined scenarios in addition to the sensitivity study presented in Section 3.3, where we identify the driving variables of the blending-distance. We aim to address this comment in the revised manuscript, either in a new subsection or as an Appendix. We are considering performing one or more new experiments where we vary a combination of at least 2 of the variables discussed in section 3.3.
- The Referee comments that it is misleading to refer to turbulent fluctuations as noise in an observations and we fully agree. That was one of the messages in Section 3.1, but it was not properly worded. We can imagine that the magnitude of these turbulent fluctuations and the impact entrainment can have close to the surface might come as a surprise for the ammonia measurement community. One of the advantages of an LES is the ability to visualize this turbulent mixing. In Section 3.1 (Fig. 2) we wanted to show that fluctuations as a result of turbulent mixing can have a large amplitude (>4 ppb) and be quite long lasting (up to 5 minutes), even in the background ammonia concentration. We will take a close look at the wording of this section and are considering removing the sentence mentioning instrumental noise to make sure our message cannot be misinterpreted.
- We are not sure if we correctly understand the comment of the Referee about Line 235. The fluctuation intensity (fI) is calculated for all concentrations, but we subtract a moving average concentration to remove the trend from the ammonia concentration. Next, we compare the fI of the total NH3 (background + plume) to the fI of the background NH3 by calculating the difference in percentages. The blending-distance is calculated by calculating the maximum distance at which the percentage changein fI (PCfI in Fig. 4) reaches a fixed threshold, e.g. a 25 % increase in fI resulting from the emission plume.
We interpret this comment as Referee #2 suggesting that it would be interesting to see how the change in fI and the blending-distance relate to the mean concentration of the emission plume. In other words, how much of an increase in the NH3 concentration is found close to the emission source. We are considering to address this question with a new figure showing the concentration of the emitted NH3 (NH3, plume).Reference:
Vila-Guerau de Arellano, J., Talmon, A.M., and Builtjes, P. J.: A chemically reactive plume model for the NO-NO2-O3 system, Atmospheric Environment. Part A. General Topics, 24, 2237–2246, https://doi.org/10.1016/0960-1686(90)90255-L, 1990.
-
AC2: 'Reply on RC2', Ruben Schulte, 24 Feb 2022
Ruben B. Schulte et al.
Ruben B. Schulte et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 363 | 114 | 21 | 498 | 7 | 6 |
- HTML: 363
- PDF: 114
- XML: 21
- Total: 498
- BibTeX: 7
- EndNote: 6
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1