the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Apr 2025
Modeling impacts of ozone on gross primary production across European forest ecosystems using JULES
Abstract. This study investigates the effects of tropospheric ozone (O3), a potent greenhouse gas and air pollutant, on European forests, an issue lacking comprehensive analysis at the site level. Unlike other greenhouse gases, O3 in the troposphere is primarily formed through photochemical reactions, significantly impairing vegetation productivity and carbon fixation, thereby impacting forest health and ecosystem services. We utilise data from multiple European flux tower sites and integrate statistical and mechanistic modelling approaches to simulate O3 impacts on photosynthesis and stomatal conductance. The study examines six key forest sites across Europe: Hyytiälä and Värriö (Finland), Brasschaat (Belgium), Fontainebleau-Barbeau (France), Bosco-Fontana, and Castelporziano 2 (Italy), representing boreal, temperate, and Mediterranean climates. These sites provide a diverse range of environmental conditions and forest types, enabling a comprehensive assessment of O3 effects on Gross Primary Production (GPP). We calibrated the Joint UK Land Environment Simulator (JULES) model using observed GPP data to simulate different O3 exposure sensitivities. Incorporating O3 effects improved the model’s accuracy across all sites, although the magnitude of improvement varied depending on site-specific factors such as vegetation type, climate, and ozone exposure levels. The GPP reduction due to ozone exposure varied considerably across sites, with annual mean reductions ranging from 1.04 % at Värriö to 6.2 % at Bosco-Fontana. These findings emphasise the need to account for local environmental conditions when assessing ozone stress on forests. This study highlights the strengths and limitations of the JULES model in representing O3-vegetation interactions, providing critical insights for predicting forest health and productivity under future air pollution scenarios. The model effectively captures the diurnal and seasonal variability of GPP and its sensitivity to O3 stress, particularly in boreal and temperate forests. However, its performance is limited in Mediterranean ecosystems, where pronounced O3 peaks and environmental stressors such as high vapor pressure deficit exacerbate GPP declines, pointing to the need for improved parameterisation and representation of site-specific processes. By integrating in situ measurements, this research contributes to developing targeted strategies for mitigating the adverse effects of O3 on forest ecosystems.
Competing interests: The authors declare that they have no conflict of interest.
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.- Preprint
(1561 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2025-1375', Anonymous Referee #1, 30 Apr 2025
Dear authors,
the manuscript 'Modeling impacts of ozone on gross primary production across European forest ecosystems using JULES' highlights the need of account for O3 damage on European forests and its gross primary productivity. The JULES model shows a good performance at all six sites against GPP measurements which could be improved by optimization. While the model findings are important and the manuscript is generally well written, I unfortunately found the third research question, the most interesting one, ' how can an optimised model help us understand these (the O3 damage) mechanisms' not fully answered.
Major concerns:
In order to transfer concrete learning and recommendations for the land modelling community I propose to describe how you do the parameter optimisation. To enhance the paper's significance it would be good to also think along the lines of these questions: Do you expect these parameters also apply to other places worldwide? Are the findings of your study model-specific or what can other models learn from it in terms of model improvement/development? Also, I would like to see more concrete interpretation of the results specifying the importance of environmental stressors on stomatal conductance and the direct O3 stress. A measure of how you define a forest to be sensitive or resilient to O3 would help.
Minor concerns:
line 14/15: difficult to read, please reformulate/split.
line 28/29: 'providing critical insights for predicting forest health and productivity under future air pollution scenarios. ' What do you mean by 'critical insights'?
line 54/55: An average change cannot lead to a bigger change in a sub-region. Please correct/reformulate.
line 57: 'interactions' is quite broad. Can you be more specific here? E.g. In populated regions, O3 precursors mainly stem from traffic emissions.
Section 2.1: describing the climate zone at each site would help the analysis and interpretation of the results later.
Section 2.2: Please mention the measurement uncertainty at least of GPP and LE also in the text (e.g. in relative terms)
Fig. 2a: The blue line is hardly visible
line 160: incorporated O3 and CO2 as forcing data?
eq. 1 and 2 use different notation for multiplication
eq. 3 (not numbered): How is the wilting point soil moisture and critical soil moisture defined?
line 163: add one sentence on why the O3 damage is applied separately
line 202/203: The reader would be curious to see the specific parameters for 'a' and 'FO3,crit': mention it here, in a table in the SI or reference the source
line 219: L-BFGS-B is not defined like this anywhere
Fig. 3 is not immediately clear, the arrows could be smaller, you can give more words and more structure
line 266: 'are sensitivity' ?
line 289: With which simulation do you do the partial correlation?
line 310-312: complicated sentence , please reformulate so that is more smooth
line 332/333: Isn't O3 concentration just quite low at Hyy?
line 347 and 350: adjustments to -> adjustments of ?
line 348: so is water limitation here more important than the O3 stress?
line 354: 'the addition of O3'. Pretend that additional O3 is added as forcing to the simulation, misleading.
Section 3.2: mention the relative change in the text helps more than the absolute values and differences
line 380/81: What do you mean? VPD is an env. stress factor. High VPD would mean low stomatal opening (in most cases)
line 385-87: This statement is counteracting for me. Why do accounting of O3 effects makes such a big improvement although Hyy forest is not much sensitive to O3 stress?
line 401: mention which parameters (in brackets)
line 449/450: linking climatic variable to antioxidant production does not fit here in my opinion
Citation: https://doi.org/10.5194/egusphere-2025-1375-RC1 -
AC1: 'Reply on RC1', Inês Vieira, 03 Jul 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1375/egusphere-2025-1375-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Inês Vieira, 03 Jul 2025
-
RC2: 'Comment on egusphere-2025-1375', Anam Khan, 22 May 2025
Dear authors,
This manuscript presents a site-level analysis of tropospheric ozone (O3) impacts on gross primary productivity (GPP) at European forests covering boreal, temperate, and Mediterranean regions. Site-level O3 sensitivities are presented using a partial correlation analysis and various parameterizations of a land surface model, JULES. A comparison of model-data agreement in GPP estimates are presented between JULES parameterizations that do and do not include O3 effects. These comparisons are used to develop a case in favor of incorporating O3 effects in a land surface model and to study such effects through site-level optimizations of the model’s parameters. Overall, the manuscript is a significant contribution, and it is well-written and well-organized. Please see my comments below.
General comments
O3 effects or O3 damage factors often use PFT-specific or a single species’ response to O3 to assess regional to global impacts. Vieira et al. present an important advancement by running site-level optimizations of key land surface model parameters (including those that determine O3 damage factors) using GPP partitioned from observed net carbon flux. The authors further contribute an analysis of high O3 days and investigate the impacts separately. I find these contributions to be the most important part of this study, and research questions 1 and 2 are well-addressed. However, I find the answer to question 3 incomplete. While it is interesting to see the diurnal patterns of latent heat flux, VPD, GPP, and [O3], the interactions are not clear without the missing links: stomatal conductance and stomatal uptake of O3. I have listed a series of specific and technical comments below. I hope these comments can help clarify many of the interesting findings of this study.
Specific comments
- Since the partitioned GPP is central to the inference made in this manuscript, it would help if the authors offered a description of how GPP was partitioned from observed net carbon flux. This could be as simple as a brief description with a reference to a citation that details the methods.
- Lines 130 – 132 claim that GPP and LE were estimated from net-carbon flux. Net carbon flux is used to estimate net ecosystem exchange and GPP. LE is not estimated from net carbon flux. It is typically estimated from H2O flux. The authors should consider correcting or clarifying if they have developed a technique or used an existing technique to estimate LE from net carbon flux.
- The JULES damage scheme calculated the O3 damage factor, F, as a function of the stomatal flux of O3 (equation 7). It appears that this is the instantaneous stomatal flux of O3. However, the cumulative flux of O3 through stomata is typically used as the damaging quantity (Lombardozzi et al., 2013, Wittig et al., 2007). Many threshold-based O3 damage indicators are based on cumulative exposure or cumulative stomatal dose (i.e.: AOT40 and POD6). The authors could consider elaborating on this in the discussion section of this manuscript by discussing if it would be worthwhile to use cumulative O3 stomatal flux in future optimization studies.
- The JULES O3 damage factor, F, as it is formulated in the current study appears to be the same damage factor that is applied to both stomatal conductance (gp) and net photosynthesis (A). However, previous research suggests that net photosynthesis and stomatal conductance are differentially impacted by O3 (Lombardozzi et al., 2012a,b). Both quantities might not exhibit the same sensitivity to O3 or might not change at the same rate as a function of O3 uptake (Lombardozzi et al., 2012b). This suggests the use of separate damage factors, sensitivities, and critical O3 levels for stomatal conductance and net photosynthesis. Are a and FO3crit separately estimated for A and gp? These distinctions are important because they might have implications for modeling transpiration in a land surface model if stomatal conductance is involved. The results report only one value for FO3crit and a which implies that the same damage factor is applied to both A and gp. In the discussion portion of the paper, it would be worth discussing the reasoning behind the JULES modeling choices for the specific formulation of O3 stress on gp and A compared to other methods of incorporating damage factors in land surface models (see Lombardozzi et al. 2012a and b who tried various configurations of an O3 damage factor in the community land model).
- The diurnal cycles of partitioned and JULES simulated GPP are shown in Figure 5. Can the authors clarify whether these diurnal cycles were estimated using data and simulations from all seasons or just the summer?
- Some statements about the diurnal cycle of GPP need clarification. The authors mention midday depressions in GPP at Mediterranean sites at line 394 and again at lines 465 - 467. Can the authors specify which GPP estimates show these midday depressions (partitioned or simulated)? The partitioned GPP from flux data (black line in diurnal plots) do not show midday depressions at the Italian sites (There does appear to be somewhat of a morning depression in partitioned GPP at IT-BFt). The simulated GPP suggests midday depression and diurnal asymmetry (higher fluxes in the morning) at the IT-BFt.
- The discussion of O3 interactions with environmental factors on high ozone days (in section 3.3 and in the discussion section) needs more clarification and elaboration. It seems that the authors are using LE as a simple proxy for stomatal conductance (LE increases or decreases with changes in stomatal conductance). It could be helpful if the authors plotted the diurnal cycles of JULES simulated stomatal conductance and stomatal flux, FO3, as a third column in Fig. 7.
- At line 380, the authors mention that the midday peak of VPD and LE facilitates greater O3 uptake through higher stomatal conductance. This appears to be the case at many sites where the reduction in GPP from simulations that did not include O3 (reduction in GPP from purple line to green line) appear to be the highest during the midday period previously defined by the authors (12 – 16). However, this does not seem to be the case for IT-BFt. The largest reduction in GPP at IT-BFt during high O3 days appears to take place in the morning hours when [O3] is not at peak. It appears that LE and VPD are also high before the 12 – 16 midday period at IT-BFt. Can the authors discuss this interesting exception more? Is there high morning stomatal conductance and morning stomatal O3 flux at this site?
- The results about the boreal sites in section 3.3 can use more elaboration and clarification. Throughout the section, the authors use RMSE reductions to quantify O3 At line 382, the authors mention that O3 impacts on the boreal sites (FI-Hyy and FI-Var) are limited. However, the RMSE reductions between optimalizations with and without O3 at FI-Hyy are the largest among the sites (9.97 down to 0.52). This implies the impact of O3 peaks is the strongest at the boreal site, FI-Hyy, compared to all other sites. Can the authors clarify or limit their statement to FI-Var?
- Are the authors referring to the partial correlation analysis when saying that FI-Hyy is less sensitive to O3 overall (at line 387)? The JULES parameter optimization seems to suggest otherwise: FI-Hyy has higher sensitivity, a, and lower FO3crit among the sites (Figure 6). Is FI-Hyy less sensitive to O3 or does it receive less O3 exposure outside of select high O3 days?
- Line 385: Can the authors clarify what they mean by “simulations without O3 significantly underestimate GPP”? In Fig. 7, it appears that the simulations without O3 (purple line) estimate much higher GPP compared to the partitioned GPP (black line).
- The authors could consider revising the section on Mediterranean sites (starting at like 394). As I mentioned in the previous comment, I am particularly concerned about the claim that compared to other sites, the Italian sites exhibit stronger O3 induced reductions in GPP (line 395). Again, FI-Hyy appears to exhibit the largest reduction in RMSE during high O3 days (a reduction from 9.97 to 0.52). BE-Bra also shows a higher or comparable reduction in RMSE (7.57 down to 3.09) compared to IT-Cp2 and IT-BFt. This needs to be corrected or clarified.
- The claim at line 465 needs elaboration: “Southern sites like IT-BFt and IT-Cp2 exhibited pronounced midday declines in GPP, reflecting their heightened sensitivity to ozone and the compounding effects of high VPD and LE.”
- The model simulated midday declines in GPP only appear at IT-BFt in Fig. 7. Please clarify what the authors mean by midday (12 – 16 hour) decline in GPP at IT-Cp2.
- The authors mention compounding effects of high VPD and LE at the southern sites at line 466 attempting to make a case for multiple stressors exacerbating ozone impacts. At IT-BFt, I can see the authors’ claim in the model simulations. The modeling does suggest that GPP declines past the 10th hour when VPD is high and further declines when O3 impacts are added to the modeling. However, the partitioned GPP (black line) does not show this type of compound stress at IT-BFt. Partitioned GPP is showing the opposite. It increased into the afternoon hours (after 10 when VPD is high) which suggest there is not much midday or afternoon water stress. The authors might want to elaborate on these differences between the partitioned GPP and JULES simulated GPP when discussing the potential of a compound water stress and O3
Technical comments
- It is difficult to tell the sites apart in Fig. 2a.
- The factor 1.6 on line 168 is a factor to convert from conductance to CO2 to conductance to H2O (ratio of CO2 and H2O diffusivities). The conductance to water vapor is gp.
- Should FO3 and FO3crit be in different units in equation 7? I am looking at line 201.
- Remove second comma after “vegetation” in line 243.
- Figure 7: It might help to double-check the units for VPD on the y-axis. Is it supposed to be displayed in hPa (not kPa)?
- Consider picking a consistent way to write GPP reductions in section 3.4. The authors make it clear that negatives mean decreases and continue to use negative quantities throughout most of the section. You could consider changing 5.22% to -5.22% at line 424 for consistency.
References
Lombardozzi, D., Levis, S., Bonan, G., Sparks, J.P., 2012a. Predicting photosynthesis and transpiration responses to ozone: decoupling modeled photosynthesis and stomatal conductance. Biogeosciences 9, 3113–3130. https://doi.org/10.5194/bg-9-3113-2012
Lombardozzi, D., Sparks, J.P., Bonan, G., Levis, S., 2012b. Ozone exposure causes a decoupling of conductance and photosynthesis: implications for the Ball-Berry stomatal conductance model. Oecologia 169, 651–659. https://doi.org/10.1007/s00442-011-2242-3
Lombardozzi, D., Sparks, J.P., Bonan, G., 2013. Integrating O3 influences on terrestrial processes: photosynthetic and stomatal response data available for regional and global modeling. Biogeosciences 10, 6815–6831. https://doi.org/10.5194/bg-10-6815-2013
Wittig, V.E., Ainsworth, E.A., Long, S.P., 2007. To what extent do current and projected increases in surface ozone affect photosynthesis and stomatal conductance of trees? A meta-analytic review of the last 3 decades of experiments. Plant, Cell & Environment 30, 1150–1162. https://doi.org/10.1111/j.1365-3040.2007.01717.x
Citation: https://doi.org/10.5194/egusphere-2025-1375-RC2 -
AC2: 'Reply on RC2', Inês Vieira, 03 Jul 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1375/egusphere-2025-1375-AC2-supplement.pdf
- Since the partitioned GPP is central to the inference made in this manuscript, it would help if the authors offered a description of how GPP was partitioned from observed net carbon flux. This could be as simple as a brief description with a reference to a citation that details the methods.
-
RC3: 'Comment on egusphere-2025-1375', Anonymous Referee #3, 01 Jun 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1375/egusphere-2025-1375-RC3-supplement.pdf
-
AC3: 'Reply on RC3', Inês Vieira, 03 Jul 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1375/egusphere-2025-1375-AC3-supplement.pdf
-
AC3: 'Reply on RC3', Inês Vieira, 03 Jul 2025
Status: closed
-
RC1: 'Comment on egusphere-2025-1375', Anonymous Referee #1, 30 Apr 2025
Dear authors,
the manuscript 'Modeling impacts of ozone on gross primary production across European forest ecosystems using JULES' highlights the need of account for O3 damage on European forests and its gross primary productivity. The JULES model shows a good performance at all six sites against GPP measurements which could be improved by optimization. While the model findings are important and the manuscript is generally well written, I unfortunately found the third research question, the most interesting one, ' how can an optimised model help us understand these (the O3 damage) mechanisms' not fully answered.
Major concerns:
In order to transfer concrete learning and recommendations for the land modelling community I propose to describe how you do the parameter optimisation. To enhance the paper's significance it would be good to also think along the lines of these questions: Do you expect these parameters also apply to other places worldwide? Are the findings of your study model-specific or what can other models learn from it in terms of model improvement/development? Also, I would like to see more concrete interpretation of the results specifying the importance of environmental stressors on stomatal conductance and the direct O3 stress. A measure of how you define a forest to be sensitive or resilient to O3 would help.
Minor concerns:
line 14/15: difficult to read, please reformulate/split.
line 28/29: 'providing critical insights for predicting forest health and productivity under future air pollution scenarios. ' What do you mean by 'critical insights'?
line 54/55: An average change cannot lead to a bigger change in a sub-region. Please correct/reformulate.
line 57: 'interactions' is quite broad. Can you be more specific here? E.g. In populated regions, O3 precursors mainly stem from traffic emissions.
Section 2.1: describing the climate zone at each site would help the analysis and interpretation of the results later.
Section 2.2: Please mention the measurement uncertainty at least of GPP and LE also in the text (e.g. in relative terms)
Fig. 2a: The blue line is hardly visible
line 160: incorporated O3 and CO2 as forcing data?
eq. 1 and 2 use different notation for multiplication
eq. 3 (not numbered): How is the wilting point soil moisture and critical soil moisture defined?
line 163: add one sentence on why the O3 damage is applied separately
line 202/203: The reader would be curious to see the specific parameters for 'a' and 'FO3,crit': mention it here, in a table in the SI or reference the source
line 219: L-BFGS-B is not defined like this anywhere
Fig. 3 is not immediately clear, the arrows could be smaller, you can give more words and more structure
line 266: 'are sensitivity' ?
line 289: With which simulation do you do the partial correlation?
line 310-312: complicated sentence , please reformulate so that is more smooth
line 332/333: Isn't O3 concentration just quite low at Hyy?
line 347 and 350: adjustments to -> adjustments of ?
line 348: so is water limitation here more important than the O3 stress?
line 354: 'the addition of O3'. Pretend that additional O3 is added as forcing to the simulation, misleading.
Section 3.2: mention the relative change in the text helps more than the absolute values and differences
line 380/81: What do you mean? VPD is an env. stress factor. High VPD would mean low stomatal opening (in most cases)
line 385-87: This statement is counteracting for me. Why do accounting of O3 effects makes such a big improvement although Hyy forest is not much sensitive to O3 stress?
line 401: mention which parameters (in brackets)
line 449/450: linking climatic variable to antioxidant production does not fit here in my opinion
Citation: https://doi.org/10.5194/egusphere-2025-1375-RC1 -
AC1: 'Reply on RC1', Inês Vieira, 03 Jul 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1375/egusphere-2025-1375-AC1-supplement.pdf
-
AC1: 'Reply on RC1', Inês Vieira, 03 Jul 2025
-
RC2: 'Comment on egusphere-2025-1375', Anam Khan, 22 May 2025
Dear authors,
This manuscript presents a site-level analysis of tropospheric ozone (O3) impacts on gross primary productivity (GPP) at European forests covering boreal, temperate, and Mediterranean regions. Site-level O3 sensitivities are presented using a partial correlation analysis and various parameterizations of a land surface model, JULES. A comparison of model-data agreement in GPP estimates are presented between JULES parameterizations that do and do not include O3 effects. These comparisons are used to develop a case in favor of incorporating O3 effects in a land surface model and to study such effects through site-level optimizations of the model’s parameters. Overall, the manuscript is a significant contribution, and it is well-written and well-organized. Please see my comments below.
General comments
O3 effects or O3 damage factors often use PFT-specific or a single species’ response to O3 to assess regional to global impacts. Vieira et al. present an important advancement by running site-level optimizations of key land surface model parameters (including those that determine O3 damage factors) using GPP partitioned from observed net carbon flux. The authors further contribute an analysis of high O3 days and investigate the impacts separately. I find these contributions to be the most important part of this study, and research questions 1 and 2 are well-addressed. However, I find the answer to question 3 incomplete. While it is interesting to see the diurnal patterns of latent heat flux, VPD, GPP, and [O3], the interactions are not clear without the missing links: stomatal conductance and stomatal uptake of O3. I have listed a series of specific and technical comments below. I hope these comments can help clarify many of the interesting findings of this study.
Specific comments
- Since the partitioned GPP is central to the inference made in this manuscript, it would help if the authors offered a description of how GPP was partitioned from observed net carbon flux. This could be as simple as a brief description with a reference to a citation that details the methods.
- Lines 130 – 132 claim that GPP and LE were estimated from net-carbon flux. Net carbon flux is used to estimate net ecosystem exchange and GPP. LE is not estimated from net carbon flux. It is typically estimated from H2O flux. The authors should consider correcting or clarifying if they have developed a technique or used an existing technique to estimate LE from net carbon flux.
- The JULES damage scheme calculated the O3 damage factor, F, as a function of the stomatal flux of O3 (equation 7). It appears that this is the instantaneous stomatal flux of O3. However, the cumulative flux of O3 through stomata is typically used as the damaging quantity (Lombardozzi et al., 2013, Wittig et al., 2007). Many threshold-based O3 damage indicators are based on cumulative exposure or cumulative stomatal dose (i.e.: AOT40 and POD6). The authors could consider elaborating on this in the discussion section of this manuscript by discussing if it would be worthwhile to use cumulative O3 stomatal flux in future optimization studies.
- The JULES O3 damage factor, F, as it is formulated in the current study appears to be the same damage factor that is applied to both stomatal conductance (gp) and net photosynthesis (A). However, previous research suggests that net photosynthesis and stomatal conductance are differentially impacted by O3 (Lombardozzi et al., 2012a,b). Both quantities might not exhibit the same sensitivity to O3 or might not change at the same rate as a function of O3 uptake (Lombardozzi et al., 2012b). This suggests the use of separate damage factors, sensitivities, and critical O3 levels for stomatal conductance and net photosynthesis. Are a and FO3crit separately estimated for A and gp? These distinctions are important because they might have implications for modeling transpiration in a land surface model if stomatal conductance is involved. The results report only one value for FO3crit and a which implies that the same damage factor is applied to both A and gp. In the discussion portion of the paper, it would be worth discussing the reasoning behind the JULES modeling choices for the specific formulation of O3 stress on gp and A compared to other methods of incorporating damage factors in land surface models (see Lombardozzi et al. 2012a and b who tried various configurations of an O3 damage factor in the community land model).
- The diurnal cycles of partitioned and JULES simulated GPP are shown in Figure 5. Can the authors clarify whether these diurnal cycles were estimated using data and simulations from all seasons or just the summer?
- Some statements about the diurnal cycle of GPP need clarification. The authors mention midday depressions in GPP at Mediterranean sites at line 394 and again at lines 465 - 467. Can the authors specify which GPP estimates show these midday depressions (partitioned or simulated)? The partitioned GPP from flux data (black line in diurnal plots) do not show midday depressions at the Italian sites (There does appear to be somewhat of a morning depression in partitioned GPP at IT-BFt). The simulated GPP suggests midday depression and diurnal asymmetry (higher fluxes in the morning) at the IT-BFt.
- The discussion of O3 interactions with environmental factors on high ozone days (in section 3.3 and in the discussion section) needs more clarification and elaboration. It seems that the authors are using LE as a simple proxy for stomatal conductance (LE increases or decreases with changes in stomatal conductance). It could be helpful if the authors plotted the diurnal cycles of JULES simulated stomatal conductance and stomatal flux, FO3, as a third column in Fig. 7.
- At line 380, the authors mention that the midday peak of VPD and LE facilitates greater O3 uptake through higher stomatal conductance. This appears to be the case at many sites where the reduction in GPP from simulations that did not include O3 (reduction in GPP from purple line to green line) appear to be the highest during the midday period previously defined by the authors (12 – 16). However, this does not seem to be the case for IT-BFt. The largest reduction in GPP at IT-BFt during high O3 days appears to take place in the morning hours when [O3] is not at peak. It appears that LE and VPD are also high before the 12 – 16 midday period at IT-BFt. Can the authors discuss this interesting exception more? Is there high morning stomatal conductance and morning stomatal O3 flux at this site?
- The results about the boreal sites in section 3.3 can use more elaboration and clarification. Throughout the section, the authors use RMSE reductions to quantify O3 At line 382, the authors mention that O3 impacts on the boreal sites (FI-Hyy and FI-Var) are limited. However, the RMSE reductions between optimalizations with and without O3 at FI-Hyy are the largest among the sites (9.97 down to 0.52). This implies the impact of O3 peaks is the strongest at the boreal site, FI-Hyy, compared to all other sites. Can the authors clarify or limit their statement to FI-Var?
- Are the authors referring to the partial correlation analysis when saying that FI-Hyy is less sensitive to O3 overall (at line 387)? The JULES parameter optimization seems to suggest otherwise: FI-Hyy has higher sensitivity, a, and lower FO3crit among the sites (Figure 6). Is FI-Hyy less sensitive to O3 or does it receive less O3 exposure outside of select high O3 days?
- Line 385: Can the authors clarify what they mean by “simulations without O3 significantly underestimate GPP”? In Fig. 7, it appears that the simulations without O3 (purple line) estimate much higher GPP compared to the partitioned GPP (black line).
- The authors could consider revising the section on Mediterranean sites (starting at like 394). As I mentioned in the previous comment, I am particularly concerned about the claim that compared to other sites, the Italian sites exhibit stronger O3 induced reductions in GPP (line 395). Again, FI-Hyy appears to exhibit the largest reduction in RMSE during high O3 days (a reduction from 9.97 to 0.52). BE-Bra also shows a higher or comparable reduction in RMSE (7.57 down to 3.09) compared to IT-Cp2 and IT-BFt. This needs to be corrected or clarified.
- The claim at line 465 needs elaboration: “Southern sites like IT-BFt and IT-Cp2 exhibited pronounced midday declines in GPP, reflecting their heightened sensitivity to ozone and the compounding effects of high VPD and LE.”
- The model simulated midday declines in GPP only appear at IT-BFt in Fig. 7. Please clarify what the authors mean by midday (12 – 16 hour) decline in GPP at IT-Cp2.
- The authors mention compounding effects of high VPD and LE at the southern sites at line 466 attempting to make a case for multiple stressors exacerbating ozone impacts. At IT-BFt, I can see the authors’ claim in the model simulations. The modeling does suggest that GPP declines past the 10th hour when VPD is high and further declines when O3 impacts are added to the modeling. However, the partitioned GPP (black line) does not show this type of compound stress at IT-BFt. Partitioned GPP is showing the opposite. It increased into the afternoon hours (after 10 when VPD is high) which suggest there is not much midday or afternoon water stress. The authors might want to elaborate on these differences between the partitioned GPP and JULES simulated GPP when discussing the potential of a compound water stress and O3
Technical comments
- It is difficult to tell the sites apart in Fig. 2a.
- The factor 1.6 on line 168 is a factor to convert from conductance to CO2 to conductance to H2O (ratio of CO2 and H2O diffusivities). The conductance to water vapor is gp.
- Should FO3 and FO3crit be in different units in equation 7? I am looking at line 201.
- Remove second comma after “vegetation” in line 243.
- Figure 7: It might help to double-check the units for VPD on the y-axis. Is it supposed to be displayed in hPa (not kPa)?
- Consider picking a consistent way to write GPP reductions in section 3.4. The authors make it clear that negatives mean decreases and continue to use negative quantities throughout most of the section. You could consider changing 5.22% to -5.22% at line 424 for consistency.
References
Lombardozzi, D., Levis, S., Bonan, G., Sparks, J.P., 2012a. Predicting photosynthesis and transpiration responses to ozone: decoupling modeled photosynthesis and stomatal conductance. Biogeosciences 9, 3113–3130. https://doi.org/10.5194/bg-9-3113-2012
Lombardozzi, D., Sparks, J.P., Bonan, G., Levis, S., 2012b. Ozone exposure causes a decoupling of conductance and photosynthesis: implications for the Ball-Berry stomatal conductance model. Oecologia 169, 651–659. https://doi.org/10.1007/s00442-011-2242-3
Lombardozzi, D., Sparks, J.P., Bonan, G., 2013. Integrating O3 influences on terrestrial processes: photosynthetic and stomatal response data available for regional and global modeling. Biogeosciences 10, 6815–6831. https://doi.org/10.5194/bg-10-6815-2013
Wittig, V.E., Ainsworth, E.A., Long, S.P., 2007. To what extent do current and projected increases in surface ozone affect photosynthesis and stomatal conductance of trees? A meta-analytic review of the last 3 decades of experiments. Plant, Cell & Environment 30, 1150–1162. https://doi.org/10.1111/j.1365-3040.2007.01717.x
Citation: https://doi.org/10.5194/egusphere-2025-1375-RC2 -
AC2: 'Reply on RC2', Inês Vieira, 03 Jul 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1375/egusphere-2025-1375-AC2-supplement.pdf
- Since the partitioned GPP is central to the inference made in this manuscript, it would help if the authors offered a description of how GPP was partitioned from observed net carbon flux. This could be as simple as a brief description with a reference to a citation that details the methods.
-
RC3: 'Comment on egusphere-2025-1375', Anonymous Referee #3, 01 Jun 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1375/egusphere-2025-1375-RC3-supplement.pdf
-
AC3: 'Reply on RC3', Inês Vieira, 03 Jul 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2025/egusphere-2025-1375/egusphere-2025-1375-AC3-supplement.pdf
-
AC3: 'Reply on RC3', Inês Vieira, 03 Jul 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 631 | 89 | 24 | 744 | 18 | 40 |
- HTML: 631
- PDF: 89
- XML: 24
- Total: 744
- BibTeX: 18
- EndNote: 40
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1