This manuscript describes an initial evaluation of how possible future tree planting (up to the potential carrying capacity) would interact with climate change to change water availability. The study is a useful first estimate of an important practical problem since tree planting is often promoted as being a suitable mitigation strategy. The topic is suitable for the journal and the paper is very well written and easy to follow. I had very few substantive comments. As an initial study of a complex phenomenon it has many drawbacks but to the credit of the authors many of those are described in detail in section 4.

We thank Professor Michael Roderick for the positive feedback on the paper and constructive comments to improve the paper. We reply in detail to the comments below.

Methods:

• One aspect of the project design/methodology that was not commented upon is the inclusion of the Budyko process to separate P into Q and ET. I did not understand why this step was included since the climate model projections include ET and Q and you could use those directly. (See Roderick et al 2015; Milly & Dunne 2017). (Also see point 3 below.)
  
  The climate model projections indeed include ET and Q but retrieving these directly from the climate model projections would not allow us to show the exact effects of the tree cover change on the ET and Q fluxes. With the Budyko calculation approach we can first calculate the ET and Q fluxes with a ‘basis’ tree cover (used in scenario present climate and scenario climate change) and thereafter we calculate the ET and Q fluxes with a new tree cover (used in scenario climate change and tree cover change). The difference in ET and Q fluxes under climate change is then ascribed to the changing tree cover. With this approach we assume a consistent tree cover (per scenario) for the five CMIP6 climate models in this study and thus that we utilize the exact same tree cover change and the same ET methodology in our calculations for each of the five models and each of the scenarios. 
• I think it might be useful to extend the discussion a little further to give full context. The idea of planting trees to their “potential” and then assessing the hydrologic changes is challenging but useful. Overall the global impacts of tree planting (or climate change) on runoff were small (Fig. 3) but could be important locally. Imagine we actually did plant all of those trees. The other impacts are on biodiversity (likely increase) but the big impact would be on agricultural production which would decrease (by a lot). Something should be said about this in the discussion.
  
  Our study is hypothetical, but planting all of those trees could indeed have important impacts on biodiversity and agricultural production. In chapter 3.2 (Spatial impact of climate change and future tree cover change on hydrological fluxes) we will describe that, while global effects are small, these effects can be significant locally. We will also mention that our scenario is hypothetical, and that it can have far-reaching effects on agricultural production, 

Details:

1. Line 16. “…. Water availability with more than 5 mm yr-1 .” I did not understand what the 5 mm yr-1 was referring to?
   
   This sentence refers to the information provided in Fig. 6 of the manuscript, whereby ‘more than 5 mm yr-1’ indicates the extent of water availability decrease as a result of climate change and tree cover change. In other words, this sentence explains that a part of the land surface (14 %) could locally experience relatively large decreases in water availability. We suggest to change line 16 for more clarity into; 
   
   ‘However, for 14 % of the land surface, both tree cover change and climate change could decrease water availability by more than 5 mm yr-1.’
2. Line 34. Typo. … increased by more
   
   Thank you, corrected.
3. Line 403-404. Yes, the Budyko models do not explicitly account for that but the climate models do (see Yang et al 2019, Hydrologic implications of vegetation response to elevated CO2 in climate projections, Nature Climate Change). This begs the question of why use the Budyko models to split the P between ET and Q. You could use climate model output directly and avoid the Budyko step in the methodology.
   
   As mentioned above, we indeed use the Budyko models as a simplified approach which allows us to adjust the tree cover and thereafter assess the exact impact on the water fluxes as a result of changing the tree cover. Retrieving the water fluxes directly from the climate models for scenario ‘climate change + tree cover’ would require us to run those climate models with an altered tree cover. 

This study investigates the hydrological impacts of large-scale tree cover changes and climate changes using the Budyko framework with data from CMIP6 models, tree cover datasets, and the UTrack moisture recycling dataset. The authors claim that this study provides a first estimate of the impacts of tree cover change under climate change on global hydrological fluxes over land. However, the novelty of the methods and conclusions is significantly overstated. Numerous studies have already addressed this topic, and the methodological flaws prevent me from recommending this paper for publication at its current stage. 

We thank reviewer 2 for their feedback comments on the paper. We address the comments in detail below.

My major concerns are as follows:

• The study estimates ET and Q using the Budyko equations, assigning fixed parameters for scenarios with and without tree cover. This approach predetermines the impact of tree cover change on the water balance, reducing the subsequent calculations to a numerical exercise with limited scientific value. Moreover, the implicit assumption behind this method—that changes in Budyko parameters are solely related to tree cover—is problematic. Climate factors, such as snow proportion and rainfall intensity, also influence Budyko parameters and affect the estimates of ET and Q. At a minimum, the authors should validate their approach by demonstrating its ability to estimate ET and Q using observational data.

We agree that the Budyko parameters are also affected by other factors besides vegetation, such as soil water characteristics (Gunkel and Lange, 2017), or snowfall conditions (Berghuijs et al., 2014). Our assumption that the changes in Budyko parameters are solely related to tree cover is therefore indeed simplified, however, we disagree that this is problematic for our study. Several studies (e.g. Zhang et al., 2001) have shown that vegetation type explains a large part of the variability in rainfall-runoff ratios on a multi-annual scale, indicating that Budyko models can be applied to study the impact of vegetation on runoff. By using six different Budyko models, we provided insight into the uncertainty of the Budyko calculations, which is (among others) related to the fraction of variability in runoff that is not explained by vegetation. 

Furthermore, we do not validate the research approach in our manuscript since this methodology was already tested by Hoek van Dijke et al. (2022). In that study, the mean streamflow (Q) obtained with the Budyko calculations for a present climate was validated with observational data for 19 large river basins (see Fig. 1b in Hoek van Dijke et al. (2022)). We will refer to this methodology validation by Hoek van Dijke et al. (2022) in the revised version of the manuscript.

The study of Hoek van Dijke et al. (2022) applied Budyko calculations and the UTrack moisture tracking dataset to analyse the influence of large-scale potential tree cover change on hydrological fluxes. Their results generally align with the findings of Tuinenburg et al. (2022) who calculated the effect of potential tree cover change on present climate evaporation with a global hydrological model and the UTrack moisture tracking model. Hence, all of the above added to our confidence that we could utilize the Budyko models (and the UTrack datasets) to estimate the changes in hydrological fluxes following tree cover change.

• The logic behind combining historical UTrack data with projected future precipitation to represent the scenario of ‘Climate change with tree cover change with altered moisture recycling’ is highly confusing. These datasets cover different time periods, and the authors themselves admit that "The UTrack dataset does not account for the feedbacks of altered tree cover." Additionally, forest-climate feedbacks are complex and can affect net radiation through changes in albedo, surface temperature, and other factors, thereby influencing PET. Why do the authors focus solely on precipitation changes without considering the potential impacts on the energy balance?

We understand that our use of the historical UTrack moisture tracking dataset for a future climate may seem confusing. However, as also mentioned in lines 189 - 191 of the manuscript, there is not yet a global moisture tracking dataset readily available for the future time period under climate change. Therefore, the UTrack moisture tracking dataset and the climate change scenarios cover different time periods as we are limited by the availability of future moisture tracking data. This shortcoming was clearly stated in Section 2.4 and the implications were discussed in Section 4.2 of the manuscript. 

Furthermore, we agree with the reviewer that forest-climate feedbacks are complex and that our methodology does not account for feedbacks of tree cover change on the atmospheric moisture recycling and the energy balance. Including these feedbacks in the context of this study would require running the five CMIP6 models - with a coupled biosphere and atmosphere - both with and without large-scale tree cover change, which could introduce other uncertainties (discussed in lines 439 - 448 of the manuscript). We already briefly address that our approach omits the feedbacks of a changing tree cover on the energy balance through e.g. surface temperatures and albedo (lines 429 - 431). However, we will extend our discussion on the implications of leaving out the potential impacts on the energy balance. Below you can find the suggested additional text for the discussion section, starting from line 429 onwards;

Moreover, this research does not account for tree cover change feedbacks on moisture recycling via changes in e.g. the surface albedo, cloud cover, atmospheric carbon dioxide concentrations, surface temperatures, length of moisture transport pathways, and global circulation. These feedbacks can impact the energy balance, e.g. increased tree cover can lower the surface albedo and subsequently enhance the surface temperature and thereby also impact e.g., moisture recycling ratios and locations. However, these feedbacks are complex and can be contrasting, depending on the location of the landuse change, and the same holds for representing those feedbacks in earth-system models (Portmann et al., 2022; De Hertog et al., 2023; King et al., 2024). Therefore, this constrains our ability to predict how the exclusion of these feedbacks may have affected our results.'

• There is a lack of coherence between the simulation results and tree cover datasets. The authors use the global tree cover dataset provided by Hansen et al. (2013) as the tree cover data for the ‘present’ scenario, but the climate models they use for the ‘present’ scenario do not incorporate this tree cover dataset. The same issue arises for future simulations with potential tree cover changes. In conjunction with my first point, it seems that the study's methodology is more of a numerical combination of different datasets rather than a robust scientific analysis. This raises doubts about the reliability of the results.

In this research we use a consistent tree cover across all CMIP6 models for each research scenario, instead of varying the tree cover for each model. Every CMIP6 model has its own land cover map, and we are aware that the tree covers in our approach deviate from those in the CMIP6 models. We chose this method for two different reasons; 

1. The Budyko model calculations are performed for each of the research scenarios and CMIP6 models. By using a consistent tree cover dataset across the five CMIP6 models in the Budyko calculations, we exclude the impacts of model-related tree cover on the calculated hydrological fluxes and thereby minimize flux differences between the models. Hence, we aim that the variations in calculated hydrological fluxes across the CMIP6 models are solely related to differences in model climate conditions.
2. For each of the CMIP6 models, the future potential tree cover by Roebroek (2023) is based upon the tree cover by Hansen et al. (2013) instead of the tree cover corresponding to the CMIP6 model. Hence, in our research we use the tree cover by Hansen et al. (2013) as the ‘base’ present tree cover to ensure consistency between the present tree cover and the future potential tree cover. 

For the future potential tree cover we underline that the tree cover under the SSP3-7.0 pathway in CMIP6 models deviates from the future potential tree cover used in our approach. Therefore, the feedback loop between potential tree cover change and climate conditions (a changing tree cover would alter the climate conditions which in turn would change the tree cover, and so on) is not accounted for in this research. We discuss the limitations regarding the usage of the future potential tree cover map and the exclusion of feedback loops corresponding to tree cover change in lines 384-396 of the manuscript. 

Finally, we do not agree with the reviewer that our study is a ‘numerical combination of different datasets rather than a robust scientific analysis’. Numerical combinations are the basis for many scientifically robust studies. Our data-driven approach has its shortcomings, as also clearly discussed in the manuscript, but it also has its strengths and faces different sources of uncertainty than earth system model studies. For example, the Budyko models allow us to account for the feedbacks on soils, rooting systems, and litter layer following (re)forestation, whereas these processes are omitted in most earth system models. Furthermore, a recent model-based study by King et al. (2024) analyses the hydrological impacts of (re)forestation in a future climate using a single climate model, therefore not considering the biases present in global climate models. Since our study adopts a multi-model approach by including data from five CMIP6 models, our methodology allows us to account for biases in e.g., precipitation which can differ substantially per model (Fig. 2a of the manuscript). It should also be noted that the predicted effects of climate change on vegetation are highly uncertain in earth system models (Terrer et al., 2019; Roebroek et al., 2024).

Overall, we believe that our study complements earth system models studies, which encounter different sources of uncertainties, and thereby contributes to this topic from a different perspective. We clearly acknowledge the constraints of this study and present our results as a first estimate of the hydrological impacts of tree cover change in a future climate.

• The analysis is overly simplistic. While the paper devotes significant space to introducing tree cover change, it notes that tree cover change "is not used for calculations and is solely used to visualize the differences in tree cover between the present climate and CC, and the CC+TCC scenarios."

We agree with the reviewer that this sentence was a bit unclear and could be interpreted incorrectly, therefore we removed it from the revised version of the manuscript to avoid further confusion.



The final result only reiterates that while climate change increases global runoff, changing tree cover reverses this effect, leading to a limited net impact on runoff compared to the present climate and current tree cover. This conclusion has been well documented in the literature, and the paper does not offer any additional insights.

We disagree with the reviewer that our conclusion has been well documented in previous literature. There have indeed been other publications which addressed this topic, however, most of these studies were either on a different spatial scale (e.g., Buechel et al. (2024) focused on the UK), focused on one earth system model (King et al., 2024), focused on a different climate pathway (e.g., Buechel et al. 2024; King et al., 2024), or did not consider climate change (Hoek van Dijke et al., 2022). Our study provides a global overview of how large-scale tree cover change under climate pathway SSP3-7.0 can amplify, mitigate, or reverse negative effects of climate change on runoff. Hence, while the globally averaged impacts of climate and tree cover changes indeed result in a limited net change in runoff, we also show that their effects on water availability can differ substantially on a regional scale.

Finally, as mentioned before, we consider it an asset if this complex topic is studied from various research perspectives, by using different methodologies and recording the corresponding uncertainties.

• On a positive note, the authors have acknowledged many of these issues in their discussion. However, merely acknowledging these limitations is not enough to mitigate their impact on the reliability and scientific integrity of the paper. A more rigorous scientific approach is needed to explore this important and interesting topic.

There are indeed shortcomings in our study and given these constraints we present our results as a first estimate of the hydrological effects due to climate change and tree cover change. By underlining the shortcomings of our research methodology we provide future studies with potential next steps and guidelines for further improvement and validation of our findings. As mentioned before, we believe that each methodology has specific shortcomings and advantages and by following a diversity of approaches in different studies, we can generate a clear overview of this challenging topic.

References:

Berghuijs, W., Woods, R. & Hrachowitz, M.: A precipitation shift from snow towards rain leads to a decrease in streamflow, Nature Clim Change 4, 583–586, https://doi.org/10.1038/nclimate2246, 2014.

Buechel, M., Berthou, S., Slater, L., Keat, W., Lewis, H., and Dadson, S.: Hydrometeorological response to afforestation in the UK: findings from a kilometer-scale climate model, Environmental Research Letters, 19, https://doi.org/10.1088/1748-9326/ad4bf6, 2024.

Gunkel, A., & Lange, J.: Water scarcity, data scarcity and the Budyko curve—An application in the Lower Jordan River Basin, Journal of Hydrology: Regional Studies, 12, 136-149. https://doi.org/10.1016/j.ejrh.2017.04.004, 2017.

Hansen, M. C., Potapov, P. V., Moore, R., Hancher, M., Turubanova, S. A., Tyukavina, A., Thau, D., Stehman, S. V., Goetz, S. J., Loveland, 580 T. R., Kommareddy, A., Egorov, A., Chini, L., Justice, C. O., and Townshend, J. R. G.: High-Resolution Global Maps of 21st-Century Forest Cover Change, Science, 342, 850–853, https://doi.org/10.1126/science.1244693, 2013.

Hoek van Dijke, A. J., Herold, M., Mallick, K., Benedict, I., Machwitz, M., Schlerf, M., Pranindita, A., Theeuwen, J. J. E., Bastin, J.-F., and Teuling, A. J.: Shifts in regional water availability due to global tree restoration, Nature Geoscience, 15, 363–368, https://doi.org/10.1038/s41561-022-00935-0, 2022.

King, J. A., Weber, J., Lawrence, P., Roe, S., Swann, A. L. S., and Val Martin, M.: Global and regional hydrological impacts of global forest expansion, Biogeosciences, 21, 3883–3902, https://doi.org/10.5194/bg-21-3883-2024, 2024.605

Roebroek, C. T.: Exploring the limits of forest carbon storage for climate change mitigation, Doctoral thesis, ETH Zurich, https://doi.org/10.3929/ethz-b-000655156, 2023.

Roebroek, C. T. J., Caporaso, L., Alkama, R., Duveiller, G., Davin, E. L., Seneviratne, S. I., & Cescatti, A.: Climate policies for carbon neutrality should not rely on the uncertain increase of carbon stocks in existing forests, Environmental Research Letters, 19(4), https://doi.org/10.1088/1748-9326/ad34e8, 2024.

Terrer, C., Jackson, R. B., Prentice, I. C., Keenan, T. F., Kaiser, C., Vicca, S., Fisher, J. B., Reich, P. B., Stocker, B. D., Hungate, B. A., Peñuelas, J., McCallum, I., Soudzilovskaia, N. A., Cernusak, L. A., Talhelm, A. F., Van Sundert, K., Piao, S., Newton, P. C. D., Hovenden, M. J., Blumenthal, D. M., Liu, Y. Y., Müller, C., Winter, K., Field, C. B., Viechtbauer, W., Van Lissa, C. J., Hoosbeek, M. R., Watanabe, M., Koike, T., Leshyk, V. O., Polley, H. W., & Franklin, O.: Nitrogen and phosphorus constrain the CO2 fertilization of global plant biomass, Nature Climate Change, 9(9), 684-689. https://doi.org/10.1038/s41558-019-0545-2, 2019.

Tuinenburg, O. A., Bosmans, J. H. C., and Staal, A.: The global potential of forest restoration for drought mitigation, Environmental Research 685 Letters, 17, 1–8, https://doi.org/10.1088/1748-9326/ac55b8, 2022.

Zhang, L., Dawes, W. R., and Walker, G. R.: Response of mean annual evapotranspiration to vegetation changes at catchment scale, Water 710 Resources Research, 37, 701–708, https://doi.org/10.1029/2000wr900325, 2001.

We thank reviewer #3 for their review comments. Please find our detailed response in the attachment.