the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 13 Feb 2024
Can large-scale tree cover change negate climate change impacts on future water availability?
Abstract. The availability of fresh water over land may become increasingly scarce under climate change, and natural and human-induced tree cover changes can further enhance or negate the water scarcity. Previous studies showed that global tree cover change can have large impacts on water availability under current climate conditions, but did not touch upon the implications of global tree cover change under climate change. Here, we study the hydrological impacts of large-scale tree cover change (climate-induced changes in combination with large-scale afforestation) in a future climate (SSP3-7.0) following an interdisciplinary approach. By combining data from five CMIP6 climate models with a future potential tree cover dataset, six Budyko models, and the UTrack moisture recycling dataset, we can disentangle the impacts of climate change and future tree cover change on evaporation, precipitation, and runoff. We quantify per grid cell and for five selected river basins (Yukon, Mississippi, Amazon, Danube, and Murray-Darling) if tree cover changes enhance or counteract the climate-driven changes in runoff due to their impact on evapotranspiration and moisture recycling. Globally averaged, the impacts of climate change and large-scale tree cover change on runoff are of similar magnitude with opposite signs. While climate change increases the global runoff, the changing tree cover reverses this effect which overall results in a limited net impact on runoff relative to the present climate and current tree cover. Nevertheless, locally the change in runoff due to tree cover change and climate change can be substantial with increases and decreases of more than 100 mm year-1. We show that for approximately 16 % of the land surface, tree cover change can increase the water availability significantly. However, we also find that, for 14 % of the land surface, both tree cover change and climate change might decrease water availability with more than 5 mm year-1. For each of the selected catchments, the direction and magnitude of the impacts of climate change and tree cover change vary, with dominating climate change impacts in all basins except the Mississippi River basin. Our results show that ecosystem restoration projects targeting an altered tree cover should consider the corresponding hydrological impacts to limit unwanted (non-)local reductions in water availability.
- Preprint
(8647 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-313', Anonymous Referee #1, 05 Mar 2024
The study analyses the effects of climate and tree cover changes on water availability used a wide range of available data and by combining different methods. The effects of climate change and tree cover changes are distinguished from each other in order to understand the importance of each for changes in evaporation, precipitation and runoff both on a global scale and on 5 specific basins. Overall they find that climate change induced increases in runoff can be cancelled due to decreases caused by tree cover change and argue that these effects should be considered in future restoration projects. Â
I find the study topic and approaches used highly interesting and relevant, however as the authors note it is a challenging topic given the many uncertainties. The authors use a blend of different data and approaches (budyko models, moisture tracking) to analyse the effects in a consistent way using the CMIP6 climate model simulations. These results and approaches are presented clearly and are highly relevant for the study topic despite the large uncertainties of the study topic. The limitations of the study are clearly and openly described which helps a lot in guiding the interpretation of the results. Overall I believe the study proposes an interesting data based approach which is a substantial addition to present day literature. I only have a few minor comments related to the approach and interpretation of the results:
Methods:
- The used methodological workflow is creative and allows to use currently existing datasets to answer this complex problem. The authors spent a lot of time and effort in describing the limitations of this approach (such as the lack of an interactive atmosphere and thus climate feedbacks) which is important to take into account. However, one caveat in the analysis setup is barely discussed but could also be important. The authors base their analysis on CMIP6 climate model simulations and then apply the climate model data with different independent tree cover datasets. When the tree cover dataset is kept the same (e.g. both 2000 land cover) they attribute this difference in CMIP model results fully to climate change. However, this neglects the fact that each of these simulations have transient land cover related to their specific SSP scenarios as defined by the LUH2 dataset (https://luh.umd.edu/ , Hurtt et al. 2020). To my surprise this associated land cover pathway to the SSP scenarios is never mentioned despite the knowledge that those land cover changes can also cause climate effects in those simulations. The authors compare a mid-century period to a historical period so the land cover changes might be modest overall but it is currently unclear whether the authors checked this. I do not think that this caveat undermines the used approach (especially on the global scale) but I would find it important that the authors reflect upon possible biases due to this such as over areas of large land cover changes, such as in specific basins, in SSP3 as the climate change signal could also contain a land cover induced signal.
- A smaller methodological question: why did the authors apply different averaging periods for tree cover changes (2041-2060) and climate (2035-2064)? This seems strange and is not clearly explained in the methods section.
Discussion and conclusion:
- The study presents a data driven approach which despite large uncertainties and limitations in setup and available data manages to deliver plausible results which is highlighted by the agreement between changes reported in this study and more local based studies. However, it remains unclear due to the variety of limitations within the approach how these results can be interpreted. Currently it is only described as a ‘first estimate’ which remains a bit vague. I believe that some more clarification would help regarding the meaning of the effects/calculated values. In essence how can one interpret/use these results? Do these results represent an upper boundary of potential effects or rather a lower estimate due to the neglected climate feedbacks? Are these values representative and applicable for local studies (e.g. region/basin scale) or only at the global scale of the utilised scenario (i.e. idealised case)? These are not questions that the study can quantify directly but as the proposed framework is flexible and applicable with available data hence it would be interesting to have some more indication on how the neglected effects could change the overall outcome.
- line 460: It is a bit confusing what is exactly meant by local coupled model simulations. In general the text refers to the need (and lack of) coupled earth system model simulations, could you be more specific in highlighting which type of simulations are required and would help move research forward (or would be applicable for this approach). And to what extent those should be local (regional afforestation simulations maybe)?
Technical comments:
- line 109: reference got messed up
- line 254-258: good that you test for statistical significance but why only for Q and just for one scenario, why not include all models and scenarios and only show significant changes on the maps if most grid cells show significant effects in any case?
- Line 370 and 433: constrainTs
- line 339-340: this is confusing, if you want to highlight how important restoration potential is you can better report absolute values here (at least together with the relative values).
Â
Â
Citation: https://doi.org/10.5194/egusphere-2024-313-RC1 -
AC1: 'Reply on RC1', Imme Benedict, 02 May 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-313/egusphere-2024-313-AC1-supplement.pdf
-
RC2: 'Comment on egusphere-2024-313', Anonymous Referee #2, 29 Mar 2024
Introduction:
The manuscript presents an innovative analysis on how large-scale tree cover changes might influence the impacts of climate change on future water availability, focusing on both global and specific basin scales. The interdisciplinary approach, combining Budyko models, CMIP6 climate projections, tree cover datasets, and moisture tracking, offers valuable insights into a highly complex issue. Despite acknowledging the inherent uncertainties and methodological limitations, the paper contributes significantly to the ongoing discourse on climate change mitigation strategies and ecosystem restoration. The clarity in presenting the complex methodologies and the transparent discussion of limitations is notable. In addition to the first referee’s comments, to strengthen the paper further, detailed below are minor issues requiring attention.
Method:
-The paper uses a set of assumptions in its modeling approach, especially in applying the Budyko framework and the UTrack dataset for future climate scenarios. A more detailed justification of these assumptions and their validity in the context of climate change, and how they influence the results would strengthen the paper’s scientific rigor. For instance, the sensitivity of Budyko model to vegetation type and coverage. Alternatively, you could mention how these assumptions might impact the findings in light of previous research in the discussion section.
Uncertainties:
- The spatial resolution of the datasets used, including CMIP6 projections, tree cover, and moisture recycling data, may influence the study's conclusions, especially when scaling down to specific basins. The impact of using a uniform spatial resolution across diverse ecological and climatic zones could introduce inaccuracies in regions with high spatial variability in climate and land cover, which is worthy of mentioning.
- The study seems to treat tree cover changes as static between the two periods compared. However, tree cover dynamics, including growth rates, succession stages, and potential dieback could significantly affect water cycling processes. it is worth mentioning these temporal dynamics might influence the study's outcomes.
Textual corrections:
Line 55: The phrase "allows to study" could be grammatically improved to "allows us to study."
Line 290: "93mmyr−1" should include a space for clarity, "93 mm yr−1."
Citation: https://doi.org/10.5194/egusphere-2024-313-RC2 -
AC2: 'Reply on RC2', Imme Benedict, 02 May 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-313/egusphere-2024-313-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Imme Benedict, 02 May 2024
Status: closed
-
RC1: 'Comment on egusphere-2024-313', Anonymous Referee #1, 05 Mar 2024
The study analyses the effects of climate and tree cover changes on water availability used a wide range of available data and by combining different methods. The effects of climate change and tree cover changes are distinguished from each other in order to understand the importance of each for changes in evaporation, precipitation and runoff both on a global scale and on 5 specific basins. Overall they find that climate change induced increases in runoff can be cancelled due to decreases caused by tree cover change and argue that these effects should be considered in future restoration projects. Â
I find the study topic and approaches used highly interesting and relevant, however as the authors note it is a challenging topic given the many uncertainties. The authors use a blend of different data and approaches (budyko models, moisture tracking) to analyse the effects in a consistent way using the CMIP6 climate model simulations. These results and approaches are presented clearly and are highly relevant for the study topic despite the large uncertainties of the study topic. The limitations of the study are clearly and openly described which helps a lot in guiding the interpretation of the results. Overall I believe the study proposes an interesting data based approach which is a substantial addition to present day literature. I only have a few minor comments related to the approach and interpretation of the results:
Methods:
- The used methodological workflow is creative and allows to use currently existing datasets to answer this complex problem. The authors spent a lot of time and effort in describing the limitations of this approach (such as the lack of an interactive atmosphere and thus climate feedbacks) which is important to take into account. However, one caveat in the analysis setup is barely discussed but could also be important. The authors base their analysis on CMIP6 climate model simulations and then apply the climate model data with different independent tree cover datasets. When the tree cover dataset is kept the same (e.g. both 2000 land cover) they attribute this difference in CMIP model results fully to climate change. However, this neglects the fact that each of these simulations have transient land cover related to their specific SSP scenarios as defined by the LUH2 dataset (https://luh.umd.edu/ , Hurtt et al. 2020). To my surprise this associated land cover pathway to the SSP scenarios is never mentioned despite the knowledge that those land cover changes can also cause climate effects in those simulations. The authors compare a mid-century period to a historical period so the land cover changes might be modest overall but it is currently unclear whether the authors checked this. I do not think that this caveat undermines the used approach (especially on the global scale) but I would find it important that the authors reflect upon possible biases due to this such as over areas of large land cover changes, such as in specific basins, in SSP3 as the climate change signal could also contain a land cover induced signal.
- A smaller methodological question: why did the authors apply different averaging periods for tree cover changes (2041-2060) and climate (2035-2064)? This seems strange and is not clearly explained in the methods section.
Discussion and conclusion:
- The study presents a data driven approach which despite large uncertainties and limitations in setup and available data manages to deliver plausible results which is highlighted by the agreement between changes reported in this study and more local based studies. However, it remains unclear due to the variety of limitations within the approach how these results can be interpreted. Currently it is only described as a ‘first estimate’ which remains a bit vague. I believe that some more clarification would help regarding the meaning of the effects/calculated values. In essence how can one interpret/use these results? Do these results represent an upper boundary of potential effects or rather a lower estimate due to the neglected climate feedbacks? Are these values representative and applicable for local studies (e.g. region/basin scale) or only at the global scale of the utilised scenario (i.e. idealised case)? These are not questions that the study can quantify directly but as the proposed framework is flexible and applicable with available data hence it would be interesting to have some more indication on how the neglected effects could change the overall outcome.
- line 460: It is a bit confusing what is exactly meant by local coupled model simulations. In general the text refers to the need (and lack of) coupled earth system model simulations, could you be more specific in highlighting which type of simulations are required and would help move research forward (or would be applicable for this approach). And to what extent those should be local (regional afforestation simulations maybe)?
Technical comments:
- line 109: reference got messed up
- line 254-258: good that you test for statistical significance but why only for Q and just for one scenario, why not include all models and scenarios and only show significant changes on the maps if most grid cells show significant effects in any case?
- Line 370 and 433: constrainTs
- line 339-340: this is confusing, if you want to highlight how important restoration potential is you can better report absolute values here (at least together with the relative values).
Â
Â
Citation: https://doi.org/10.5194/egusphere-2024-313-RC1 -
AC1: 'Reply on RC1', Imme Benedict, 02 May 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-313/egusphere-2024-313-AC1-supplement.pdf
-
RC2: 'Comment on egusphere-2024-313', Anonymous Referee #2, 29 Mar 2024
Introduction:
The manuscript presents an innovative analysis on how large-scale tree cover changes might influence the impacts of climate change on future water availability, focusing on both global and specific basin scales. The interdisciplinary approach, combining Budyko models, CMIP6 climate projections, tree cover datasets, and moisture tracking, offers valuable insights into a highly complex issue. Despite acknowledging the inherent uncertainties and methodological limitations, the paper contributes significantly to the ongoing discourse on climate change mitigation strategies and ecosystem restoration. The clarity in presenting the complex methodologies and the transparent discussion of limitations is notable. In addition to the first referee’s comments, to strengthen the paper further, detailed below are minor issues requiring attention.
Method:
-The paper uses a set of assumptions in its modeling approach, especially in applying the Budyko framework and the UTrack dataset for future climate scenarios. A more detailed justification of these assumptions and their validity in the context of climate change, and how they influence the results would strengthen the paper’s scientific rigor. For instance, the sensitivity of Budyko model to vegetation type and coverage. Alternatively, you could mention how these assumptions might impact the findings in light of previous research in the discussion section.
Uncertainties:
- The spatial resolution of the datasets used, including CMIP6 projections, tree cover, and moisture recycling data, may influence the study's conclusions, especially when scaling down to specific basins. The impact of using a uniform spatial resolution across diverse ecological and climatic zones could introduce inaccuracies in regions with high spatial variability in climate and land cover, which is worthy of mentioning.
- The study seems to treat tree cover changes as static between the two periods compared. However, tree cover dynamics, including growth rates, succession stages, and potential dieback could significantly affect water cycling processes. it is worth mentioning these temporal dynamics might influence the study's outcomes.
Textual corrections:
Line 55: The phrase "allows to study" could be grammatically improved to "allows us to study."
Line 290: "93mmyr−1" should include a space for clarity, "93 mm yr−1."
Citation: https://doi.org/10.5194/egusphere-2024-313-RC2 -
AC2: 'Reply on RC2', Imme Benedict, 02 May 2024
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-313/egusphere-2024-313-AC2-supplement.pdf
-
AC2: 'Reply on RC2', Imme Benedict, 02 May 2024
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 522 | 156 | 34 | 712 | 20 | 18 |
- HTML: 522
- PDF: 156
- XML: 34
- Total: 712
- BibTeX: 20
- EndNote: 18
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1