the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Dec 2022
Updated Isoprene and Terpene Emission Factors for the Interactive BVOC Emission Scheme (iBVOC) in the United Kingdom Earth System Model (UKESM1.0)
Abstract. Emissions of biogenic volatile organic compounds (BVOCs) influence atmospheric composition and climate and will be influenced by future changes in land use and land cover (LULC) and change. Climate and Earth System Models typically calculate emissions using parameterisations involving surface temperature, photosynthetic activity, CO2 and the type of vegetation present. The influence of vegetation is described by assigning emission factors (EF) to different types of vegetation simulated in the model. We detail calculations of new EF for the Interactive BVOC Emission Scheme (iBVOC) used in the United Kingdom Earth System Model (UKESM). These EFs are based on those used by the Model of Emissions of Gases and Aerosols from Nature (MEGAN) v2.1 scheme.
We present these EFs as alternatives to the current EFs used in iBVOC which are derived from older versions of MEGAN and the Organizing Carbon and Hydrology in Dynamic Ecosystem (ORCHIDEE) emission scheme. The EFs currently used by iBVOC include an oversimplification which incorporates the EF of shrubs (high isoprene emitters) into the EFs for C3 and C4 grasses (low isoprene emitters) despite UKESM1 treating grasses and shrubs separately. Thus, the current approach significantly overestimates the isoprene emissions from grasses, particularly C4 grass which is responsible for 40 % of total simulated isoprene emissions in the present day, much higher than other estimates of ~0.3–10 %.
The new isoprene EF calculated in this work substantially reduce the amount of isoprene emitted by C4 grassland, in line with observational studies and other modelling approaches, while also increasing the emissions from known sources such as tropical broadleaf trees. Similar results are found from the change to terpene EF.
The total global isoprene and terpene emissions with the new EF are in the range suggested by literature. The existing model biases in isoprene column are slightly exacerbated with the new EFs although other drivers of this bias are also noted. The disaggregation of shrub and grass EFs should lead to a more faithful description of the contribution to BVOC emissions from different vegetation types, critical for understanding BVOC emissions in the pre-industrial and under different future LULC scenarios such as those including wide scale reforestation or deforestation.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1289 KB)
-
Supplement
(319 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1289 KB) - Metadata XML
-
Supplement
(319 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-748', Anonymous Referee #1, 11 Jan 2023
SUMMARY AND GENERAL REMARKS
Biogenic Volatile Organic Compounds (BVOCs) play a key role in the composition of the atmosphere. They control the oxidizing capacity (abundance of OH) and contribute substatially to the formation of secondary organic aerosol (SOA). Therefore, acurate representation of BVOC chemistry and emissions in state-of-science Earth system models is imperative.
Weber et al. present a very comprehensive study of the inteactive BVOC emission model iBVOC in UKESM1. They identify a significant shortcoming in the selection of Emission Factors by mass (EFmass) for isoprene (IEFmass) and terpenes (TEFmass) for one particular Plant Funtional Type (PFT), namely C4 grasses. This leads to a massive overprediction of isoprene and terpene emissions from this PFT, while the global total budget remains relatively uneffected (compared to the literature). So, one could say right for the wrong reasons.
Weber et al. then analyse several alternative options and approaches of deriving improved EFs and evaluate the impact of their alternative EFs on emissions and atmospheric mole fractions of isoprene and terpenes. Some sensible reccomendations for fututure direction wrap up this very useful paper.
I think that this work represents a very valuable and useful contribution to the growing literature on the UK community Earth System Model UKESM. The authors have identified a significant shortcoming in the BVOC emission model and offer reasonable and useful alternatives. They also present a very detailed analysis of their suggested changes on UKESM, albeit limited mostly to the emissions and atmospheric abundance of BVOCs themselves. This is in my view where the paper falls a bit short of its potential. BVOCs pla a key role in tropospheric ozone production, the atmospheric oxidising capacity (OH abundance), and the formation of secondary organic aerosol (SOA) with substatial impacts on the radiation budget. This is a bit of a missed opportunity I feel, although I concede freely that the amount of work required is likely massive and probaly beyond the scope of this paper. I hope that future work along the lines detailed above will follow.
With the above said I recommend the publication of this paper in GMD after the minor issues, which I will outline in the following, have been addressed.
SPECIFIC REMARKS
p1l12: "future changes in land use and land cover (LULC) and change."; redundant, please revise.
p2l50: citation "Cao et al., 2021" does not appear to be in the list of references.
p4ls99&100: please provide some refrences for JULES and UKCA.
p8l234: link; Is there perhaps another way of referencing these parameters (paper or technical documentation) that has a DOI? There is a potential risk that this link could break in future.
p11l333: reference to Table 3 should be to Table 4, I believe.
p11l349: "andorganic" should read "and organic"
p11l349: citation Mulcahy et al., 2020, doesn't appear to be in the list of references
p11l355,356: The first sentence in this paragraph seems a bit redundant, because it basically repeats what has been said in the previous paragraph.
p12l375: should read "timeseries of anthropogenic"
p12l379: citation Sellar et al., 2019, doesn't appear to be in the list of references
p13l392: should read "PFTs as discussed"
p15l366: should read "are at the lower end"
p15l367: "total emissions" is redundant
p15l498: "land use cover" --> either "land use" or "land cover" or "land use and land cover"
p16l495: "Fig2b" should be Fig3b, I believe
p16l502-504: how does this decrease of terpene emissions from needleleaf trees square with other models and observations, e.g., from Hyytialää? Haven't needle leaf trees always been thought to be important for terpene emissions at high latitudes. Please comment here and, if appropriate, in the text.
p16l508-511: how much could the bias be due to differences in climate between sim and obs? are those sims nudged? I believe they are, but maybe a cooment on that fact may help. Even if nudged, there will be significant differences in climate. Important?
p17l530: The observations in are introduced without prior discussion. No references are offered. Please add references and discuss backgound a bit.
p17l335: I am not sure I understand the argument for introducing the scaling factor here. Isoprene emission are zero at night both at the observation sites and in iBVOC (there is a diurnal cycle in the model representing real world conditions). So, why is the scaling need - and what data is scaled (I presume the reference is to the model). Please elaborate.
p19l609-611: This sentence needs discombobulation and revision.
p20: Data and Code availability. I appreciate that there are certain restrictions on the UM code and thus sharing is very complicated. However, this is not the case for all the data (and the code to process them), that heas been used in the plots and tables of this paper. All the data used to produce the plots and tablesas well as the code used to produce them must be made available publicly according to GMD policy. Ammend data availability section accordingly.
p27: reference Weber et al., 2021 --> needs to be updated to Weber et al., 2022, I believe.
Citation: https://doi.org/10.5194/egusphere-2022-748-RC1 -
RC2: 'Comment on egusphere-2022-748', Anonymous Referee #2, 10 Feb 2023
The authors have worked from different perspectives to improve the estimation of BVOC emissions in UKESM1. They mainly accomplished this by integrating values, land cover fractions, and SLW from various sources into the model. The authors also compared the modeled isoprene column with satellite data, and compared modelled emission from C4 grass with field data.
This effort is likely seen as a noteworthy accomplishment by the UKESM community, as it improves the accuracy of emission magnitudes and responses to afforestation. However, further evaluation of the data sources used in the model and the model outputs are VERY necessary as the biases are even bigger after all these changes made in the model. Then, the authors' decision to scale certain variables, such as temperature, and not others raises questions. What scientific questions are they trying to answer, and how has this influenced their decision to only scale temperature and not other variables?
The method description needs a lot of clarity, since as I read it now, I have difficulties to understand the reasons to compare so many different setups in the model. Why can't you evaluate these different dataset first and then only take the best one into your model. Then in the Maxforest scenario, the authors have employed highly intricate methods in an attempt to isolate interactions between multiple factors by creating artificial scenarios. However, the current presentation of these methods can be challenging to comprehend, making it difficult for readers like myself to fully understand the information. A clearer text with the addition of a schematic diagram would greatly enhance the readability and overall understanding of these scenarios.
Detailed comments:
The readability of the abstract needs to improve. By reading the current version of the abstract without reading the main text yet, I cannot understand the motivation to do these changes in the model, what “interactive” BVOC means, how these new EFs are calculated and why these are linked to LULC?
L58-62, It will be interesting for readers to know what types/magnitudes of impacts were found in Weber et al., 2022 study.
L64-66, it is not clear. Many models, such as MEGAN, do simulate emissions’ dependencies on meteorology, CO2 and land surface cover. I did not understand this point! So what is lacking in the model at this stage?
L81, so why does C4 have such a high emission factor in iBVOC then?
L173-175, I think it is more justifiable if there are studies that actually evaluate the performance of MEGAN-modelled emissions for grass- or shrub-dominated areas, rather than directly using their values to your model because they have separated these PFTs. It might be the case that these EF values in MEGAN were assigned without ground measurements. By adding these ‘unvalidated’ EF could lead to adding another layer of uncertainties to your model.
L182: what are the sources of these PFT distribution maps used for CESM and UKESM1?
L223-224: the degrees of impacts from different landcover datasets should be dependent on the area of study and also the differences in emission features of these PFTs. I wonder if the authors have actually tested the impacts from these different landcover datasets and if not, it needs justification.
L239-241, What are scaled and unscaled approaches? No explanation before.
Equation 4. I would imagine it should be more correct to directly calculate the emission factor of MEGAN at 30 celsius, and instead of getting a temperature scaling based on the temperature response of your model. As what has been implemented in the model now based on Equation 4, it forces MEGAN to follow the temperature response curves in your model. I don't think it makes sense.
L270-271: why are terpene emissions from broadleaf deciduous trees PAR-independent? Please clarify this!
L282-283 Why PAR-dependent should correct for temperature response?
L290: MEGAN did not consider photosynthesis.
L289: it needs some descriptions about the differences in CO2 inhibition between iBVOC and MEGAN. And how about the light response?
L394-395: what do you mean by using “PD emission” for this Maxforest scenario. So BVOC emissions were also kept in the present day conditions? Or do you only mean other gas emissions?
L396: I don’t understand this paragraph. Why set CO2 to be a constant? What do you mean “CO2 is not emitting”, how could these influence the CO2 inhibition effects on isoprene and monoterpene emissions? Is there a run where you have all these interactions turned on? I understand the authors want to isolate the impacts from LULC only, but different environmental conditions, such as CO2, at present and future can influence/determine the magnitudes of impacts from LULC.
About evaluation dataset:
L438-439: It needs to be described what the CrIS isoprene columns provide and how this data has been used for evaluation. Which runs were used to compare with the CrIS data.
L485-487: do you know why they differ so largely in terms of the coverage of evergreen tropical trees? Which one is closer to the observation-based landcover maps?
L517-519: As CrIS data is the only global datasets that are used for evaluation, it is difficult to argue that the changes on emission factors are better if the biases are increased. Please clarify.
L561: What do you mean that CLM5 SLW approach can capture the SLW of the MEGANv2.1.
L583: the authors never mention how the light responses differ between iBVOCs and MEGAN. How valid is your method that only focuses on temperature scaling, but not the others, like CO2 or light?
Citation: https://doi.org/10.5194/egusphere-2022-748-RC2 - AC1: 'Response to reviewer comments', James Weber, 30 Mar 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-748', Anonymous Referee #1, 11 Jan 2023
SUMMARY AND GENERAL REMARKS
Biogenic Volatile Organic Compounds (BVOCs) play a key role in the composition of the atmosphere. They control the oxidizing capacity (abundance of OH) and contribute substatially to the formation of secondary organic aerosol (SOA). Therefore, acurate representation of BVOC chemistry and emissions in state-of-science Earth system models is imperative.
Weber et al. present a very comprehensive study of the inteactive BVOC emission model iBVOC in UKESM1. They identify a significant shortcoming in the selection of Emission Factors by mass (EFmass) for isoprene (IEFmass) and terpenes (TEFmass) for one particular Plant Funtional Type (PFT), namely C4 grasses. This leads to a massive overprediction of isoprene and terpene emissions from this PFT, while the global total budget remains relatively uneffected (compared to the literature). So, one could say right for the wrong reasons.
Weber et al. then analyse several alternative options and approaches of deriving improved EFs and evaluate the impact of their alternative EFs on emissions and atmospheric mole fractions of isoprene and terpenes. Some sensible reccomendations for fututure direction wrap up this very useful paper.
I think that this work represents a very valuable and useful contribution to the growing literature on the UK community Earth System Model UKESM. The authors have identified a significant shortcoming in the BVOC emission model and offer reasonable and useful alternatives. They also present a very detailed analysis of their suggested changes on UKESM, albeit limited mostly to the emissions and atmospheric abundance of BVOCs themselves. This is in my view where the paper falls a bit short of its potential. BVOCs pla a key role in tropospheric ozone production, the atmospheric oxidising capacity (OH abundance), and the formation of secondary organic aerosol (SOA) with substatial impacts on the radiation budget. This is a bit of a missed opportunity I feel, although I concede freely that the amount of work required is likely massive and probaly beyond the scope of this paper. I hope that future work along the lines detailed above will follow.
With the above said I recommend the publication of this paper in GMD after the minor issues, which I will outline in the following, have been addressed.
SPECIFIC REMARKS
p1l12: "future changes in land use and land cover (LULC) and change."; redundant, please revise.
p2l50: citation "Cao et al., 2021" does not appear to be in the list of references.
p4ls99&100: please provide some refrences for JULES and UKCA.
p8l234: link; Is there perhaps another way of referencing these parameters (paper or technical documentation) that has a DOI? There is a potential risk that this link could break in future.
p11l333: reference to Table 3 should be to Table 4, I believe.
p11l349: "andorganic" should read "and organic"
p11l349: citation Mulcahy et al., 2020, doesn't appear to be in the list of references
p11l355,356: The first sentence in this paragraph seems a bit redundant, because it basically repeats what has been said in the previous paragraph.
p12l375: should read "timeseries of anthropogenic"
p12l379: citation Sellar et al., 2019, doesn't appear to be in the list of references
p13l392: should read "PFTs as discussed"
p15l366: should read "are at the lower end"
p15l367: "total emissions" is redundant
p15l498: "land use cover" --> either "land use" or "land cover" or "land use and land cover"
p16l495: "Fig2b" should be Fig3b, I believe
p16l502-504: how does this decrease of terpene emissions from needleleaf trees square with other models and observations, e.g., from Hyytialää? Haven't needle leaf trees always been thought to be important for terpene emissions at high latitudes. Please comment here and, if appropriate, in the text.
p16l508-511: how much could the bias be due to differences in climate between sim and obs? are those sims nudged? I believe they are, but maybe a cooment on that fact may help. Even if nudged, there will be significant differences in climate. Important?
p17l530: The observations in are introduced without prior discussion. No references are offered. Please add references and discuss backgound a bit.
p17l335: I am not sure I understand the argument for introducing the scaling factor here. Isoprene emission are zero at night both at the observation sites and in iBVOC (there is a diurnal cycle in the model representing real world conditions). So, why is the scaling need - and what data is scaled (I presume the reference is to the model). Please elaborate.
p19l609-611: This sentence needs discombobulation and revision.
p20: Data and Code availability. I appreciate that there are certain restrictions on the UM code and thus sharing is very complicated. However, this is not the case for all the data (and the code to process them), that heas been used in the plots and tables of this paper. All the data used to produce the plots and tablesas well as the code used to produce them must be made available publicly according to GMD policy. Ammend data availability section accordingly.
p27: reference Weber et al., 2021 --> needs to be updated to Weber et al., 2022, I believe.
Citation: https://doi.org/10.5194/egusphere-2022-748-RC1 -
RC2: 'Comment on egusphere-2022-748', Anonymous Referee #2, 10 Feb 2023
The authors have worked from different perspectives to improve the estimation of BVOC emissions in UKESM1. They mainly accomplished this by integrating values, land cover fractions, and SLW from various sources into the model. The authors also compared the modeled isoprene column with satellite data, and compared modelled emission from C4 grass with field data.
This effort is likely seen as a noteworthy accomplishment by the UKESM community, as it improves the accuracy of emission magnitudes and responses to afforestation. However, further evaluation of the data sources used in the model and the model outputs are VERY necessary as the biases are even bigger after all these changes made in the model. Then, the authors' decision to scale certain variables, such as temperature, and not others raises questions. What scientific questions are they trying to answer, and how has this influenced their decision to only scale temperature and not other variables?
The method description needs a lot of clarity, since as I read it now, I have difficulties to understand the reasons to compare so many different setups in the model. Why can't you evaluate these different dataset first and then only take the best one into your model. Then in the Maxforest scenario, the authors have employed highly intricate methods in an attempt to isolate interactions between multiple factors by creating artificial scenarios. However, the current presentation of these methods can be challenging to comprehend, making it difficult for readers like myself to fully understand the information. A clearer text with the addition of a schematic diagram would greatly enhance the readability and overall understanding of these scenarios.
Detailed comments:
The readability of the abstract needs to improve. By reading the current version of the abstract without reading the main text yet, I cannot understand the motivation to do these changes in the model, what “interactive” BVOC means, how these new EFs are calculated and why these are linked to LULC?
L58-62, It will be interesting for readers to know what types/magnitudes of impacts were found in Weber et al., 2022 study.
L64-66, it is not clear. Many models, such as MEGAN, do simulate emissions’ dependencies on meteorology, CO2 and land surface cover. I did not understand this point! So what is lacking in the model at this stage?
L81, so why does C4 have such a high emission factor in iBVOC then?
L173-175, I think it is more justifiable if there are studies that actually evaluate the performance of MEGAN-modelled emissions for grass- or shrub-dominated areas, rather than directly using their values to your model because they have separated these PFTs. It might be the case that these EF values in MEGAN were assigned without ground measurements. By adding these ‘unvalidated’ EF could lead to adding another layer of uncertainties to your model.
L182: what are the sources of these PFT distribution maps used for CESM and UKESM1?
L223-224: the degrees of impacts from different landcover datasets should be dependent on the area of study and also the differences in emission features of these PFTs. I wonder if the authors have actually tested the impacts from these different landcover datasets and if not, it needs justification.
L239-241, What are scaled and unscaled approaches? No explanation before.
Equation 4. I would imagine it should be more correct to directly calculate the emission factor of MEGAN at 30 celsius, and instead of getting a temperature scaling based on the temperature response of your model. As what has been implemented in the model now based on Equation 4, it forces MEGAN to follow the temperature response curves in your model. I don't think it makes sense.
L270-271: why are terpene emissions from broadleaf deciduous trees PAR-independent? Please clarify this!
L282-283 Why PAR-dependent should correct for temperature response?
L290: MEGAN did not consider photosynthesis.
L289: it needs some descriptions about the differences in CO2 inhibition between iBVOC and MEGAN. And how about the light response?
L394-395: what do you mean by using “PD emission” for this Maxforest scenario. So BVOC emissions were also kept in the present day conditions? Or do you only mean other gas emissions?
L396: I don’t understand this paragraph. Why set CO2 to be a constant? What do you mean “CO2 is not emitting”, how could these influence the CO2 inhibition effects on isoprene and monoterpene emissions? Is there a run where you have all these interactions turned on? I understand the authors want to isolate the impacts from LULC only, but different environmental conditions, such as CO2, at present and future can influence/determine the magnitudes of impacts from LULC.
About evaluation dataset:
L438-439: It needs to be described what the CrIS isoprene columns provide and how this data has been used for evaluation. Which runs were used to compare with the CrIS data.
L485-487: do you know why they differ so largely in terms of the coverage of evergreen tropical trees? Which one is closer to the observation-based landcover maps?
L517-519: As CrIS data is the only global datasets that are used for evaluation, it is difficult to argue that the changes on emission factors are better if the biases are increased. Please clarify.
L561: What do you mean that CLM5 SLW approach can capture the SLW of the MEGANv2.1.
L583: the authors never mention how the light responses differ between iBVOCs and MEGAN. How valid is your method that only focuses on temperature scaling, but not the others, like CO2 or light?
Citation: https://doi.org/10.5194/egusphere-2022-748-RC2 - AC1: 'Response to reviewer comments', James Weber, 30 Mar 2023
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 382 | 109 | 12 | 503 | 37 | 2 | 3 |
- HTML: 382
- PDF: 109
- XML: 12
- Total: 503
- Supplement: 37
- BibTeX: 2
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Cited
James A. King
Katerina Sindelarova
Maria Val Martin
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1289 KB) - Metadata XML
-
Supplement
(319 KB) - BibTeX
- EndNote
- Final revised paper