the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Sep 2025
Unravelling groundwater's role in soil-plant-atmosphere continuum: Integrated ecohydrological modelling approach using STEMMUS-SCOPE and MODFLOW 6
Abstract. Soil-plant-atmosphere continuum (SPAC) models are commonly used to investigate various components’ role(s) in ecosystem functioning. Yet, in most SPAC models, groundwater is ignored or at best represented in an over-simplified manner, leading to misunderstanding of its critical role in simulating soil-vegetation dynamics. This study investigates the groundwater’s role in soil-plant-atmosphere processes. To this end, an integrated ecohydrological modelling (IEM) framework is developed by coupling the STEMMUS-SCOPE SPAC model to the MODFLOW 6 integrated hydrological model. The standalone STEMMUS-SCOPE (ST-SC) and coupled STEMMUS-SCOPE-MODFLOW 6 (ST-SC-MF6) models were applied over an 8-year period (1 April 2016 – 31 March 2024) to three sites in the Netherlands (Loobos, Cabauw and Veenkampen). Simulated various essential variables, including soil moisture (θ), soil temperature (Ts), groundwater level, groundwater temperature, evapotranspiration (ET), gross primary productivity, net ecosystem exchange (NEE), and sun-induced chlorophyll fluorescence (SIF) were then compared to corresponding in-situ observations to evaluate the ST-SC and ST-SC-MF6 setups. Results indicated that the groundwater contribution is spatially and temporally variable. ST-SC-MF6 showed better agreement with observations than ST-SC for: a) Ts, and ET at Loobos, b) θ, ET, NEE, and SIF at Cabauw, and c) θ, and ET at Veenkampen. Notably, benefits of ST-SC-MF6 simulation were particularly prominent during dry periods, when shallow groundwater mitigated vegetative water stress. Overall, the proposed ST-SC-MF6 IEM helped to: (1) incorporate groundwater as a key component in the water, energy and carbon cycles, and (2) define the important role groundwater dynamics play in soil-plant-atmosphere continuum for deepening our understanding of ecosystem functioning.
Competing interests: One of the co-authors (Zhongbo Su) is a member of the editorial board of the journal. Other 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
(2933 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- CC1: 'Comment on egusphere-2025-4179', Giacomo Medici, 29 Oct 2025
-
RC1: 'Comment on egusphere-2025-4179', Emmanuel Dubois, 03 Nov 2025
The manuscript “Unravelling groundwater’s role in soil-plant-atmosphere continuum: Integrated ecohydrological modelling approach using STEMMUS-SCOPE and MODFLOW 6” presents the development of a coupled soil–plant–atmosphere model linked with a groundwater model to continuously simulate water and energy balances throughout the atmosphere–plant–soil continuum, ultimately aiming to investigate the contribution of groundwater resources to ecosystem dynamics. The topic is timely, promising, and highly relevant to the HESS audience.
Although the subject is compelling and the developed model produces promising results, the manuscript raises several major concerns:
- Significant work is needed to realign the research questions, methods, results, and interpretations throughout the manuscript. In its current form, it primarily presents the development of a new (and very interesting) tool and some results, without fully demonstrating its potential or emphasizing its potentially many scientific contributions. Section-specific comments are detailed afterwards.
- The presentation of the conceptual model raises a fundamental question regarding the representation and exchange of water fluxes between the two models, which could affect the simulated energy budget (see detailed comments).
- A major methodological limitation that undermines the entire study is the reliance on manual calibration.
- The results section could be substantially strengthened by exploring the model outputs in greater depth and focusing on the novel insights made possible by the coupling, rather than primarily presenting results that could be achieved with a fully coupled surface–subsurface model.
- The discussion section currently focuses on linking previously presented results and does not meet the expectations of a scientific discussion in HESS, which should highlight the main contributions of the study and position them within the broader state of knowledge.
Summary and short summary
The abstract is generally well written; however, the study goal (L31-32) is not addressed in the abstract or the manuscript itself. This reflects a broader issue observed throughout the paper. In addition, the main contributions of this work should be more clearly emphasized in the summary.
In the short summary, it would also be more appropriate to refer to the types of models used rather than citing their specific names, as this would make the text more accessible to a broader audience.
Introduction
The introduction is too long and difficult to follow, as its logical flow requires significant improvement. Rewriting it would help provide a clearer identification of the research gap and the goal of the study. Some comments that could assist in reworking it are provided below.
The first paragraph (L56-73) of the introduction is somewhat dangling. It begins with a discussion of the water, energy, and carbon cycles in ecosystems, then moves to essential climate variables and recent efforts to incorporate new variables to better characterize ecosystems, yet without establishing a clear connection between these ideas. The paragraph ends abruptly, without synthesizing its points or explaining their relevance to the present study.
The second paragraph (L74-89) provides a relatively good overview of the different soil-plant-atmosphere continuum models but could better articulate why explicitly representing groundwater processes in these models is important. Numerous studies have shown that vegetation dynamics cannot be accurately explained without considering groundwater interactions; this point is only introduced later in the third paragraph (L90-106), which is more general in scope and lacks a clear link to the implications for the present study. Similarly, the fifth paragraph (L117-138) returns to the absence of hydrogeological processes in soil-plant-atmosphere continuum models models, but again without a strong connection to the study’s objectives.
In the fifth paragraph (L139-162), the inclusion of a table is interesting, but it would fit more appropriately in the discussion section, where the contributions of the model developed in this study could be compared to previous work. In the introduction, it would be more effective to summarize the main limitations of existing integrated ecohydrological models to clearly highlight the knowledge gap that this study seeks to address.
Table 1: It would be useful to include information on whether each model offers an option for automatic calibration, as this would strengthen the comparative value of the table.
L165-167: The model’s goal seems overly broad, as it could apply to any integrated ecohydrological model. It would be helpful to specify more precisely what distinguishes this model or what particular integration it aims to achieve.
Paragraphs L173-184 and L185-192: These paragraphs appear more appropriate for the methods section. A brief sentence in the introduction would be enough to mention both models, especially since they are already referred to earlier in the text.
Method section
Figure 1 and subsection 2.1
Although the STEMMUS-SCOPE model explicitly accounts for root groundwater uptake, this variable is not directly transferred to Modflow. Instead, it is combined with unsaturated zone water fluxes to calculate the net groundwater recharge, which is the variable passed to Modflow. What is the advantage of grouping these two fluxes together, when they can occur at different time scales and result from independent processes? In dry periods or arid environments, when recharge is null and vegetation depends mainly on groundwater, does this mean that the recharge passed to Modflow becomes negative? How is this handled in the model? Is there a risk of misrepresenting key processes by coupling and simplifying the exchanges between models, especially when moving to 3D modelling? The question of groundwater uptake and its feedback on groundwater levels remains largely unresolved, and it would be unfortunate for a new tool such as the one developed in this study to be unable to characterize it in detail. Furthermore, could this approach bias the energy module of the coupling, since the volume of recharge influencing groundwater temperature in the model would differ from the actual recharge volume? While grouping recharge and vegetation groundwater uptake fluxes might be acceptable from a quantitative standpoint, given that they are opposite fluxes, it does not seem valid from an energy standpoint, as their respective impacts are not cumulative.
Subsection 2.2, figure 2
It appears that groundwater levels were measured only at the forested site, while no observation wells are reported at the meadow sites (Cabauw and VeenKampen). How, then, were groundwater levels determined at these sites? It would be useful to clarify this point in the text (L340-343 p.17).
It would also be important to briefly summarize the type of evapotranspiration data used, since it is later mentioned as a calibration variable. Is this evapotranspiration calculated using the Penman-Monteith method? If so, it represents potential evapotranspiration, and the authors should justify why it can be considered representative of actual evapotranspiration at the sites. If instead it is based on in-situ measurements, the measurement technique should be clearly specified.
More broadly, this raises questions about the motivation behind the selection of these sites, which appear to be pre-existing experimental locations. A short summary of previous research conducted at these sites would help readers better understand the overall context of the study.
Table 2
The MAQ dataset is not explicitly introduced earlier in the text. Providing a short presentation of it, either in the main text or at least in the table caption, would improve clarity. Based on the text (L355-356 p.18), it appears that groundwater levels were only partially available at one site, while Table 2 presents three different datasets for the sites. This should be made consistent.
It would also be helpful to include the discretization of the model layers (L372-382 p.20) in Table 2. Doing so would make it easier to compare the sites and would improve the readability of the text.
Subsection 2.3.3
It is somewhat disappointing to read that calibration was performed manually. Calibration represents a major step in both the modeling process and model development. A model that lacks an automatic calibration option is considerably less appealing for reuse, and the absence of automatic calibration results here greatly limits the scope of the study. This is particularly regrettable given that sufficient datasets are available to enable a multi-objective automatic calibration. I strongly recommend implementing such a procedure, as it would increase confidence in the subsequent results and enhance the credibility and applicability of the model. Since other coupled models of this kind already exist (Table 1), including an automatic calibration process would clearly strengthen the contribution of this work and highlight its potential for broader use.
Section 3
Presenting the results by site makes the section excessively long (subsections 3.1 to 3.3; p.22-32) and prevents any meaningful comparison between the three sites. It would be much clearer if the results were organized thematically—for example, by calibration results, plant processes, unsaturated zone, and saturated zone. Restructuring Section 3 in this way would likely allow for a deeper analysis of the simulation outcomes and better highlight the value of the proposed model coupling. It would be surprising if differences in model performance were not observed across the sites, given variations in calibration and validation data, dominant land cover, and hydrogeological context. Are there any patterns/trends in the input or calibration datasets that were successfully reproduced in the simulations? The exploration of dynamics during dry periods is particularly promising and could be developed further. How did the model represent the effects of drier conditions on vegetation? How were energy and water fluxes influenced within the atmosphere–plant–soil–groundwater continuum? What were the corresponding impacts on groundwater temperature, and how long did these effects persist? Including details on wet periods could also reveal interesting complementary insights.
This section reads as relatively superficial in comparison to the potential of the results and would definitely gain from a major restructuring and re-organizing.
For all figures in section 3: the use of red and green lines might make them difficult to interpret for color-blind readers.
Section 4
The first two paragraphs of Section 4 (L570-588; p.32-33) read as a summary of the methods and results rather than as part of the discussion. They could be rephrased and more appropriately integrated into the recommendations subsection where they could emphasize the methodological innovations developed in this study.
Paragraph L590-611 (p.33-34) is rather long, and its first half reads more like a general explanation of soil moisture dynamics than an interpretation of the study’s results. The specific contributions of this work should be clearly highlighted, as they are not currently acknowledged. The same remark applies to the rest of subsection 4.1. If this subsection were renamed “Contributions of the ST-SC-MF6 IEM to the simulation of soil moisture,” how would the paragraph be reworked to better emphasize the scientific advances achieved in this study? Comparisons with previous research are also needed: what has been done before, and how does the use of the ST-SC-MF6 IEM generate new insights or improve the representativity of the simulations?
Subsection 4.1 is titled “Effect of groundwater on the soil profile,” while 4.2 is titled “Effect of groundwater on water fluxes.” However, soil moisture and soil water potential are closely linked to water fluxes in the unsaturated zone. Should these two subsections be merged? It also seems that subsection 4.2 partly repeats the content of 4.1.
L644-658: what does the “enhancement” of flux simulation refer to (L647, L653)? Is it an increase of simulation quality? Is it the characterization of the contribution of a source to a flux?
L644-658: what does the “enhancement” of flux simulation refer to (L647, L653)? Does it mean an improvement in the quality of the simulation? or a characterization of the contributions of different sources to the simulated fluxes?
Subsections 4.1 and 4.2 require significant revisions, as they do not clearly identify the main contributions of the study. Instead, they read as dense descriptions linking simulation results together without sufficient interpretation. These sections would become much clearer if the results were presented earlier in a more structured way that guides readers toward the key insights. Moreover, it remains unclear from these two discussion subsections what specific advantages the integrated ecohydrological model offers compared with fully coupled groundwater models, which were also designed to represent surface, unsaturated, and saturated water fluxes and might perform equally well for this purpose.
Subsection 4.3 presents interesting links and could be expanded further. Could the simulation results be used to quantify the relative proportions of water sources used by plants? This analysis could be added to the results section and discussed here, allowing comparison with plant water source estimations based on isotopic methods. Another avenue worth exploring would be to assess the role of plant water uptake in reducing groundwater recharge or contributing to groundwater level depletion, alongside evaluating the influence of groundwater on sun-induced chlorophyll fluorescence.
Subsection 4.4 largely repeats limitations already discussed elsewhere in the manuscript. Instead, a concise summary of the main contributions of this work would help readers assess the novelty and scientific value of the study, as well as the potential applications of the developed model. A conceptual diagram could also be used to clarify the study’s main contributions and illustrate how the model can support other research.
Finally, it is disappointing that the temperature module of the simulations was not further explored or discussed in this section. The introduction emphased on this component and justified the development of a new integrated ecohydrological model partly on that basis. It would therefore be appropriate to examine and discuss this aspect in greater detail here.
Conclusion
L703: The term “qualitative” seems inappropriate, as the study does not include any qualitative approaches.
The conclusion is too long and overly dense, likely because it attempts to summarize all the study’s findings. While some of these findings are indeed interesting, their demonstration is not always clearly supported by the results and discussion sections.
Citation: https://doi.org/10.5194/egusphere-2025-4179-RC1 -
RC2: 'Comment on egusphere-2025-4179', Anonymous Referee #2, 23 Dec 2025
Full Title:
Unravelling groundwater’s role in soil-plant-atmosphere continuum: Integrated ecohydrological modelling approach using STEMMUS-SCOPE and MODFLOW 6
Overall evaluation：
This study investigates the role of groundwater in the soil-plant-atmosphere continuum by coupling the STEMMUS-SCOPE SPAC model with MODFLOW 6 to simulate key variables such as soil moisture, temperature, groundwater levels, and ecosystem processes at three locations in the Netherlands. While the research addresses an important topic in ecohydrological modeling, there are notable limitations that need addressing：1）although the coupled model aims to improve accuracy in vegetation carbon dynamics through groundwater integration, indicators like GPP show no significant improvement compared to single-model results, raising questions about the necessity of using a coupled approach. 2）the one-dimensional nature of the model limits its ability to capture lateral runoff dynamics, particularly in agricultural landscapes where water level fluctuations are significant. 3）It is important to improve result presentation, as current figures lack clarity. Overall, while the study has potential significance, it requires substantial revisions to address these issues and better communicate findings.
Major concerns:
- It is widely recognized that explicitly representing groundwater mass and energy within an IEM framework improves the quantification of the SPAC. A more carefully formulated hypothesis could better inspire readers.
- Three field sites and corresponding observational data were used to apply and test the coupled model. However, the rationale for selecting these specific sites and the underlying motivations for conducting the modeling work there are not detailed.
- While the effects of groundwater are noted, the discussion regarding its influence on the soil profile, water, and carbon fluxes could be expanded. Providing specific results and illustrative figures would greatly clarify this crucial aspect, which is fundamental to readers' interpretation of the work.
- The manuscript does not detail how soil heterogeneity in the profiles was addressed in the modeling. Clarifying how this critical factor is accounted for is essential for assessing the model's applicability and accuracy.
- The study's primary focus appears ambiguous: is it to present a new modeling framework, or to elucidate the specific role of groundwater within the SPAC system? Clarifying this objective would strengthen the paper's contribution. If the former, the model structure requires a more detailed and systematic introduction. If the latter, the analysis of groundwater's effects should be substantially deepened and discussed with greater insight.
Detailed comments:
- The choice of the three field sites (Loobos, Cabauw, Veenkampen) lacks justification. Merely noting that results differ across sites is insufficient; the significance of these differences for the study's conclusions remains unclear. Explaining why these sites were chosen and what their comparative results demonstrate would greatly enhance the impact of the work.
- While it is true that groundwater contributions are inherently variable, presenting this as a key finding is too general. The abstract could be more compelling by specifying how the variability manifests in your study systems (e.g., its primary drivers or magnitude) and what new understanding it provides.
- The statement attributes the benefits during dry periods to groundwater mitigating water stress. To substantiate this key conclusion, it is necessary to clarify: how was this causal relationship established? What analysis directly connects the simulated benefits to the reduction of vegetative water stress?
- Lines 41-42: The performance description here would be more concrete if supported by a specific numerical value, such as a key performance indicator or an error metric.
- The summary should move beyond stating the known importance of groundwater. Its stronger point is the demonstrated need to include groundwater processes in models, and this focus should be sharpened.
- The descriptions of the various model types are somewhat repetitive. Condensing this section would improve clarity and focus.
- Lines 61-62: The jump to ECV indicators here is logically abrupt. Consider combining this sentence with the next and adding a brief transition.
- Lines 81-85: Presenting model developments requires more than an isolated list of additions. A concise review of the baseline SPAC model’s history and key features is needed to root these advancements in a clear foundation.
- Lines 94-96: The causal claim here is unsupported. An increase in groundwater discharge does not logically equate to being “exposed” to climate change impacts. Please rephrase to accurately describe the relationship.
- Lines 100–103: The supporting role of groundwater in vegetation physiology is attributed primarily to its relative hydrological stability, not to a lag phenomenon. The wording should be adjusted to reflect this correct causal mechanism.
- Lines 110–111: This sentence appears logically disconnected from both the preceding and following content. Its relevance to the narrative flow should be clarified or it should be rephrased to establish a clearer connection.
- Lines 157–158: The statement notes that existing models “all lack the representation of groundwater heat processes and groundwater temperature simulation.” To strengthen the manuscript, please clarify whether your study addresses these specific gaps. If it does, highlighting this contribution in the abstract would underscore its significance. If not, consider phrasing the limitation in a more measured, forward-looking manner (e.g., “few studies have incorporated…”).
- Lines 165–172: This paragraph does not effectively articulate or foreground the key research contributions of this study. It should be revised to clearly underscore the novel points being presented.
- Lines 169–170: The parenthetical remark “(which will be extended to 2D/3D in a follow-up study)” pertains to future work rather than the present study’s contributions. Unless critical for contextualizing current limitations, it can be removed to maintain focus on the reported research.
- Lines 173–192: This discussion would be more logically positioned within the earlier section that introduces the first two model types, thereby improving the overall structural coherence.
- A clear scientific hypothesis is needed in the Introduction to establish focus and testability.
- Study Sites and Data: The rationale for selecting these three specific locations to apply the model requires a more objective and well-reasoned explanation. Please clarify the scientific basis for their selection (e.g., representing a gradient of key conditions, data availability, or specific hypotheses to be tested) to justify the study design.
- Lines 368–371: The content of this paragraph is more closely aligned with the site or data description. For better organizational flow, it is recommended to relocate it to Section 2.1.
- Lines 426–427: The criteria for deeming the listed RMSE values as “acceptable” are not self-evident. Please provide a justification or cite relevant references that establish these specific thresholds as benchmarks for model performance acceptance in similar studies. This is essential for assessing the rigor of your model calibration.
- The results section contains excessive descriptive text that merely reiterates what is already visible in the figures. To strengthen this section, focus on concisely presenting the key findings and their interpretation, rather than describing the data trends in detail.
- Lines 460-463: The results indicate that the model coupling did not improve the accuracy of GPP simulations. This appears to contradict the central hypothesis regarding groundwater's significant influence on vegetation. Please discuss this discrepancy. Does it suggest the influence is negligible under these conditions, or are there other explanatory factors (e.g., model limitations, data scale)?
- Model Applicability Concern: The study sites at Cabauw and Veenkampen are situated in agricultural settings. A key limitation is that the employed 1D model framework cannot account for lateral water fluxes, which are likely significant due to irrigation and pumping practices. Please address the potential impact of this simplification on the results and discuss its validity for these sites.
- The section would benefit from a more substantive and critical analysis. Currently, it tends to restate results rather than interpret their significance. For instance, the claim that comparing ST-SC and ST-SC-MF6 results “highlights the benefits” (Lines 586-588) is vague. The discussion should explicitly articulate what these specific benefits are—quantitative improvements, mechanistic insights gained, or limitations addressed—and explain why they matter in the broader context of SPAC modeling.
- The discussion appears somewhat isolated from the wider body of literature. To strengthen its scholarly contribution, findings should be actively compared and contrasted with existing studies.
- The current recommendations are overly general. To be more useful, this section should transition from stating ambitions (e.g., expanding to 2D/3D) to providing actionable insights based on this study. Please address: 1) What was the actual simulation quality at the three sites? Under what conditions did the coupled model perform well or poorly? 2)
- Based on your experience, what are the main practical or theoretical constraints (e.g., computational cost, data requirements, specific site conditions) that currently limit the model's application? 3) Is the coupled framework readily operable and transferable to other regions? What are the key prerequisites (data, expertise) and potential pitfalls for potential users?
Comments on the figures:
- Figures 3-9: Please reconsider the use of red-green color combinations, as they are not accessible to readers with color vision deficiencies. Employing a colorblind-friendly palette (e.g., blue-orange) or differentiating lines with symbols/patterns is strongly recommended.
- Figures 4, 5, 6, and 9: The key comparisons between models are difficult to discern due to high data density and overlapping traces in the main panels. To improve clarity, consider 1) plotting the difference between models directly, 2) using faceted subplots for major variables, or 3) highlighting specific time periods of interest.
Citation: https://doi.org/10.5194/egusphere-2025-4179-RC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 989 | 128 | 31 | 1,148 | 18 | 20 |
- HTML: 989
- PDF: 128
- XML: 31
- Total: 1,148
- BibTeX: 18
- EndNote: 20
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comment
Very good research on surface/groundwater interaction with a critical zone angle! Please, follow my guidance to improve the manuscript, and fix the issues
Specific comment
Lines 121-124. “Models typically do not account for heterogeneity in any of the aquifer properties (porous media or fractured-rock systems), which can be significant in shaping key groundwater processes such as recharge, capillary rise, and groundwater evapo-transpiration”. Insert recent review papers on the role of flow heterogeneities in porous and fractured aquifers in processes that involve the recharge:
- Zhang, H., Xu, Y., Kanyerere, T. (2020) A review of the managed aquifer recharge: Historical development, current situation and perspectives. Physics and Chemistry of the Earth, Parts A/B/C, 118, 102887.
- Agbotui, P. Y., Firouzbehi, F., Medici, G. (2025) Review of effective porosity in sandstone aquifers: insights for representation of contaminant transport. Sustainability, 17, 6469.
Line 192. Clearly state the aim of your hydrological research.
Line 192. Describe the 3 to 4 specific objectives of your research by using numbers (e.g., i, ii, and iii).
Line 195. Provide basic description of the climate and the lithology of the region where you have applied your method. Add just a bit more.
Line 196. You should refer to the key USGS document for MODFLOW 6.
Line 443. KGE values not wonderful. Please, justify.
Lines 695-725. Try to expand the implications of your research in the field of hydrology.
Figures and tables
Figure 1. Conceptual scheme on the left; Do you need a spatial scale?
Figure 1. There are two different figures. These figures should be 1 and 2.
Figure 9. Based on that, consider to re-scale figures 3 to 9.
Figures 3, 4, 5, 6, 8, and 9. There is room to make the graphs slightly larger. You gain in readability.