the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Apr 2021
The influence of hyporheic fluxes on regional groundwater discharge zones
Abstract. The importance of hyporheic water fluxes induced by hydromorphologic processes at the streambed scale and their consequential effects on stream ecohydrology have recently received much attention. However, the role of hyporheic water fluxes in regional groundwater discharge is still not entirely understood. Streambed-induced flows not only affect mass and heat transport in streams but are also important for the retention of solute contamination originating from deep in the subsurface, such as naturally occurring solutes as well as leakage from the future geological disposal of nuclear waste. Here, we applied a multiscale modeling approach to investigate the effect of hyporheic fluxes on regional groundwater discharge in the Krycklan catchment, located in a boreal landscape in Sweden. Regional groundwater modeling was conducted using COMSOL Multiphysics constrained by observed or modeled representations of the catchment infiltration and geological properties, reflecting heterogeneities within the subsurface domain. Furthermore, streambed-scale modeling was performed using an exact spectral solution of the hydraulic head applicable to streaming water over a fluctuating streambed topography. By comparing the flow fields of watershed-scale groundwater discharge with and without consideration of streambed-induced hyporheic flows, we found that the flow trajectories and the distribution of the travel times of groundwater were substantially influenced by the presence of hyporheic fluxes near the streambed surface. One implication of hyporheic flows is that the groundwater flow paths contract near the streambed interface, thus fragmenting the coherent areas of groundwater upwelling and resulting in narrow “pinholes” of groundwater discharge points.
- Preprint
(1783 KB) - Metadata XML
-
Supplement
(729 KB) - BibTeX
- EndNote
Status: closed
-
CC1: 'Comment on hess-2021-148', Ulrik Kautsky, 17 May 2021
The paper presents an interesting model study concerning potential effects of the hyporheic zone on groundwater flow patterns using a selection of the hydrological data available for Krycklan. In its present state the study is theoretical. The study would be strengthened if it included an analysis describing the extent to which the regional model was able to capture the hydrology at Krycklan. This would help to address questions about the validity of the regional hydrological model and may help frame results from the hyporheic, local-scale simulations as applicable to a real-world scenario.
The paper would be further strengthened if it discussed the robustness of the results as affected by the parameterizations and structural assumptions made in the numerical and conceptual models. For example, the potential effects of landscape topography, time-variation of streambed topography, rock fractures, soil stratification, spatial heterogeneity of streambed sediments, parameterization of the infiltration rate at the surface, and model boundary conditions are not discussed. Without such a discussion it cannot be deduced the extent to which variabilities in these, and other, variables will affect the hyporheic phenomena postulated in the modelling exercise; it is therefore not possible to assess the extent to which the results and/or methodologies presented in the study are relevant to other sites.
Citation: https://doi.org/10.5194/hess-2021-148-CC1 -
RC2: 'Review of article', Anonymous Referee #2, 29 May 2021
Review Hess hyporheic fluxes
General discussion of the content of the paper “The influence of hyporheic fluxes on regional groundwater discharge zones”
The paper "The influence of hyporheic fluxes on regional groundwater discharge zones" submitted to HESS by Mojarrad et al. investigates different scale flow systems by means of topographically induced groundwater circulation and the role of hyporheic flow on groundwater circulation in the river valleys. The flow systems are discussed in relation to possible accumulation of radioactive waste originating from deeper aquifer regions within the Quaternary deposits; by differentiating between re- and discharge zones.
I think that the approach, of the influence of topography on the distribution of hydraulic potential in different scales is interesting. The extent to which the link of deep groundwater circulation with the potential field of the hyporheic interstitial can be easily connected would have to be discussed in more depth, especially by delving into the processes that can also influence the potential fields at different scales.
The concept underlying the work is not new and it is surprising that the authors do not mention the fundamental work of J. Toth.
The authors cite the work of W. Zijl, who, through a Fourier analysis and consideration of the anisotropy of the geologic sequences, or important boundaries, such as the topography of the bedrock surface, or different structural properties of the bedrock, determine a series of flow cells. A reduction to two or three as proposed in this paper is probably too simplified. Thus, I would find it justified for the authors to emphasize the conceptual aspect of the work and to discuss the role of influence of topography of the river bed (i.e., the river corridor concept of Stanford and Ward), the influence of heterogeneity of the geological sequence, the role of the relative water flux contributions etc.).
Based on the work (concept) of Toth, it is not surprising that the drainage system can correspond to the exfiltration zones.
However, there are other aspects to be considered:
The River corridor concept of Stanford and Ward, shows that the topography of the riverbed and the topography of an impounding layer (can be e.g. glacial deposits, bedrock surface or discontinuities of the gradient of the riverbed along a river course) produce infiltration and exfiltration zones within the hyporheic interstitial, which can be very dominant with respect to water fluxes.
I do not know the glacial sediments of the area, but I suspect that because of the diversity of processes, heterogeneities in these deposits also lead to vertical hydraulic gradients that could significantly affect the simple potential distribution.
The structural heterogeneity of the bedrock and the character of the hydraulically relevant structures of the subsurface (shear zones, fracture patterns, etc.) can also significantly affect the anisotropy of the hydraulic conductivities, so that over long geologic times the pattern of the exfiltration zones is also not necessarily uniformly distributed along the drainage system.
The concept depth decaying hydraulic conductivity has to be approved by regional specific data. Other, more complex heterogeneity development with depth could have a strong influence on the Potential distribution.
A much more important point influencing the regional flow systems over longer times periods (transport of radionuclides from certain depths of the bedrock) is the dynamics of the development of the topography. Please discuss, how topography was shaped during the last 15’000 years. Over long time periods, topography cannot be assumed to constant.
I think the paper could benefit from a more in-depth discussion. In its current form, the argumentation is a bit too simplistic.
Some general comments:
- Although the character of the model is mostly conceptual the authors state that results from intense field investigations exist. Nevertheless, nearly none of these existing data are specified or used for calibration and/or validation of the model (i.e. character of the hyporheic zone, heterogeneity character of the glacial deposits.
-For seven different soil types in the Krycklan catchment hydraulic conductivity was obtained. A sensitivity analysis for hydraulic conductivity of streambed sediment would be interesting.
- Provide a more quantitative of visualization of discharge locations (Fig. 1), e.g. by means of point densities. Likewise, an illustration with the "pinholes" of groundwater discharge and/or "nested" flow cells (maybe for a zoom) would also help to understand the different flow processes.
- How was the catchment area delineated (surface?). And is this approach appropriate when evaluating the deep aquifer, when the shape of the topography changed over the last 15’000 years? A 3D visualization would help to better understand the geological settings in relation to the topography.
- Some repetitions could be avoided, like e.g. Software use.
Some specific comments:
- 29: Definition of "long-term"
- 54: This strongly depends on the geology which means that without a regional geological model a detailed statement due to deep groundwater discharge zones remain fragmentary
- 66: please add References of S. Todd
- 84: the history of the hydrology at different time scales will also influence the contribution of water from different flow systems in the exfiltration zones.
- 82: I cannot find information on geological heterogeneity
- 86: Darcy's law should be known to the readership?
- 103: Missing specific information on the time scale and the rates of considered processes
- 118: There should be drillings for geological information’s. A corresponding map and a stratigraphic-lithologic overview, allowing to evaluate the degree of vertical variability of hydraulic properties is missing.
- 122: a sedimentological decription of the glacial sediments would allow a better info on heterogeneity, till compris a lot of different glacial sediment types
- 134: There is no information about the geometric extension of the model and why was Comsol chosen for such a simple model of the geological setting and not modflow, feflow or any groundwater affine software?
- 137: The presented material does not allow this simplification, which project data support the statements of the cited publications?
ions
l.158: The information of the mesh sizes should be expanded with the number and delaunay quality of elements
- 175: DT has not been introduced
- 178: give a reference to the particle tracing routine used
l 181: One of the described aims in the introduction is the influence of radionuclide spreading for humans and biota. With a residence time of 320 million years it would be hard to make any predictions at all? A cumulative curve from particles reaching the surface over time could be more detailed.
At the long time scale the topography has changed and as a result the flow field also would have changed,
- 219-220: Specify "other models and statistical uncertainties"
- 239: Correct "Monte Carla"
- 335: Can you better explain the role of the Froude number on the dynamic head component?
- 348-350: Why you mention this? Is it not expected?
- 495: In the long time, changes of the Surface and streambed morphology is expected, therefore it is not clear how these changes interfere with the flow field in the deeper deposits (Quaternary and bedrock).
L 510 and following lines: Die you take into account structural aspects of the heterogeneity in the bedrock, such as heterogeneity due to shear zones , mylonites etc.
Figures:
Integrate Figures 1 & 2
4: Up-date: Northing? Points instead of stars?
Figure 9: Correct catchemnt
Citation: https://doi.org/10.5194/hess-2021-148-RC2 -
AC2: 'Response to the comments of Referee #2', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC2-supplement.pdf
-
AC1: 'Response to the Community Comments #1', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC1-supplement.pdf
-
RC2: 'Review of article', Anonymous Referee #2, 29 May 2021
-
RC1: 'Comment on hess-2021-148', Anonymous Referee #1, 20 May 2021
General comments
This paper investigates how hyporheic flow redistributes regional discharge at the scale of riverbeds in a case study and discusses the implications for the fate and transport of radionuclides from deep nuclear waste depositories. The study involves a multiscale flow and transport modelling framework and a rather sophisticated analysis of model outputs.
The main result (fragmentation of the regional discharge by hyporheic flow) is rather intuitive and could already be anticipated by looking at Tóth (1963) flow fields. This takes nothing away from the merit of the study, which achieved a proper demonstration and quantification of this phenomenon in a realistic example. Therefore, I recommend publication without any doubt.
The main issue I found is that the paper lacks a number of explanations (see below), but I am sure the authors can improve on this aspect.
Detailed comments
L51-56: I am not sure that velocity is the most relevant indicator for what you are trying to convey here. In fact, all the flow paths end up having similar velocities when approaching discharge (Cardenas and Jiang, 2010; Zijl, 1999; Zlotnik et al., 2011). Instead, I would think that the ratio of hyporheic flow (i.e., its total flow rate) to that of regional flow is more relevant in this discussion.
L59: I guess you mean “principal effects” and not “principle effects”.
L62-64: The sentence is grammatically incorrect.
L110: This section should be better put before the description of the models.
L111-114: A situation map would be useful (I would suggest including Figure S1 here).
L129: This sentence is unclear. I guess you mean: the mean annual runoff estimated from the stream discharge measurements was set as the infiltration (please correct if needed).
L129: Is it reasonable to neglect overland flow in your study area (you may be overestimating infiltration)?
L129: Can you indicate the calculated infiltration rate?
L161-162: I suggest referring to Haitjema and Mitchell-Bruker (2005) in support of this sentence.
L162-164: I suggest referring to Bresciani et al. (2016a, 2016b) in support of this sentence.
L166-173: This only makes sense if hydraulic head is specified and equal to the topography along the top boundary, but you just said above that you are using a recharge condition, so I am lost here.
L134-185: What are the horizontal limits of the domain?
L194: I do not understand the meaning of “Cdamp(λi)”. Is Cdamp a function of λi (I would think not since Cdamp seems to be treated as a constant)? And if it is, shouldn’t it be λij?...
L206-216: How does this relate to the previous paragraph?
L187-216: What are the extent and boundary conditions of the hyporheic flow model? I guess the boundary conditions must be head = 0 everywhere but the top boundary so as to keep a continuous solution when doing the superposition...?
L219: The term “models” is confusing here. I guess you refer to the other parameters of the hyporheic flow model and all the parameters of the catchment-scale model.
L239: “Carlo”, not “Carla”.
L247: What are “the” cubes? You have not talked about cubes before.
L252: What does “the 1552” refer to?
L253-256: How did you distinguish between intermediate and deep groundwater flow paths from these particles (did you track them backward as well)? Furthermore, how can you be sure that some of these particles are not hyporheic flow?
L264-272: This part could be clearer. Did you focus on the same 1552 areas as above (I guess so)? How did you determine the coherent catchment-scale discharge areas (I guess this would involve particle tracking and a certain grouping method)?
L308-310: I think the differences between the three layers are mostly independent of the hierarchical structure of flow cells (which was not evaluated in this study, by the way).
L331: Define the Froude number.
L418-420: So is it a good news (less exposure time of aquatic sediments)?
L421: Why would it lead to higher exposure if the exposure time is shorter?
References
Bresciani, E., Gleeson, T., Goderniaux, P., de Dreuzy, J.R., Werner, A.D., Wörman, A., Zijl, W., Batelaan, O., 2016a. Groundwater flow systems theory: Research challenges beyond the specified-head top boundary condition. Hydrogeol. J. 24, 1087–1090. https://doi.org/10.1007/s10040-016-1397-8
Bresciani, E., Goderniaux, P., Batelaan, O., 2016b. Hydrogeological controls of water table-land surface interactions. Geophys. Res. Lett. 43, 9653–9661. https://doi.org/10.1002/2016gl070618
Cardenas, M.B., Jiang, X.-W., 2010. Groundwater flow, transport, and residence times through topography-driven basins with exponentially decreasing permeability and porosity. Water Resour. Res. 46, W11538. https://doi.org/10.1029/2010wr009370
Haitjema, H.M., Mitchell-Bruker, S., 2005. Are water tables a subdued replica of the topography? Ground Water 43, 781–786. https://doi.org/10.1111/j.1745-6584.2005.00090.x
Tóth, J., 1963. A theoretical analysis of groundwater flow in small drainage basins. J. Geophys. Res. 68, 4795–4812. https://doi.org/10.1029/JZ068i016p04795
Zijl, W., 1999. Scale aspects of groundwater flow and transport systems. Hydrogeol. J. 7, 139–150. https://doi.org/10.1007/s100400050185
Zlotnik, V.A., Cardenas, M.B., Toundykov, D., 2011. Effects of multiscale anisotropy on basin and hyporheic groundwater flow. Ground Water 49, 576–583. https://doi.org/10.1111/j.1745-6584.2010.00775.x
Citation: https://doi.org/10.5194/hess-2021-148-RC1 -
AC3: 'Response to the comments of Referee #1', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC3-supplement.pdf
-
AC3: 'Response to the comments of Referee #1', Brian Babak Mojarrad, 10 Jul 2021
-
CC2: 'Comment on hess-2021-148', Ulrik Kautsky, 31 May 2021
We had an additional important comment to the manuscript lost in my previous post
Many of the claims regarding ecology and human health are outside of the scope of the study especially given the simplifications that have been made regarding the hydrology of the biosphere. An attempt to extrapolate results into these areas is not properly motivated. Results of the study will stand for themselves without having to extrapolate them into areas outside of the study’s scope which, at present, is largely theoretical.
Citation: https://doi.org/10.5194/hess-2021-148-CC2 -
AC4: 'Response to the Community Comments #2', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC4-supplement.pdf
-
AC4: 'Response to the Community Comments #2', Brian Babak Mojarrad, 10 Jul 2021
-
RC3: 'Comment on hess-2021-148', Anonymous Referee #3, 31 May 2021
General
The authors of this study used a complex multiscale model to compare the travel time and spatial distribution of catchment scale groundwater flow with and without the influence of hyporheic flow. The results are loosely linked to the fate of nuclear particles that might potentially leak out from nuclear waste disposal in the bedrock of the catchment. The main finding of the study is the fragmentation of the groundwater discharge areas, the reduction of the effective volume that carries groundwater flow and the consequential reduction of the groundwater travel time in the upper 5m of the domain. A monte carlo approach was used to some extend to generate a distribution of possible outcomes where exact model parameters were unknown.
The underlying model is without a question quite sophisticated. However, I wonder if the right model was chosen to answer the specific scientific question. The main finding, that the discharge zone is fragmented, is relatively obvious and has been published various times in model studies from the hyporheic perspective (e.g. Boano et al 2008, Trauth et al 2014, Fox et al 2016) and also investigated experimentally (e.g. Bhaskar 2012). Given that the general effect is well known, I would have expected an in-depth analysis of the correlations of at least some of the parameters involved. I wonder if this complex, multiscale, pseudo-realistic model is the right choice to draw systematic conclusions with results that have a direct use for other scientists. The model has too many unknowns to investigate general findings, e.g. correlating discharge area reduction to the ratio of hyporheic head and groundwater head or something similar.
If the authors didn’t aim to draw general results but rather calculate the effect for a specific case (nuclear waste disposal in this specific catchment), I see two problems. First, case studies do not match the scope of the HESS journal. Second, some of the boundary conditions have been selected without verification or are too unclear to be used for a specific use case (especially a safety relevant topic like nuclear waste disposal). Generally, I think that a lot of choices for boundary conditions are justified only by references to past work of the own workgroup. In some cases, references to independent workgroups would strengthen the trustworthiness of the model (see detailed comments below).
The second major finding is the reduced travel time within the upper 5m of the 500m deep domain. Also here, I miss a systematic investigation of the underlying mechanism. E.g. what is the correlation between the fragmentation and the travel time? What is the correlation with the hyporheic head and groundwater head? Is the conductivity a relevant factor in this correlation? These kind of questions should be answered with specific results rather than vague discussions, because the vague answers to these questions are obvious. Again, it would have been much easier to answer these kind of questions if the model was less complex.
In addition to the missing universal results, I don’t understand why the reduction of travel time in the upper 5m is deemed relevant. The effect of the hyporheic zone decreases exponentially with depth (see eq(5) or Elliot & Brooks 1997a), which is why only the very last segment of a particles travel path will be influenced. In Line 353 it says that the effect of the hyporheic head is strongest, where the overall seepage velocity is low. That means that particles that traveled centuries to millennia (Fig 5) to reach the surface will lose a few years on their last few meters (Fig 7). Even when the different retention coefficients in bedrock and deposits were considered (missing in Fig 7), the effect should be minor. I miss a clear explanation what processes are potentially influenced by the change in travel time in this last stage of the streampath.
Overall, I’m afraid that the study design is not able to answer the research questions to a degree that goes beyond the intuitive and well-published findings.
Specific:
102: I don’t think it is necessary to show the definition of Darcy’s velocity, but if you do, please use the already defined symbol for Darcy’s velocity “W_c” instead of q and q_seepage.
108: The 50-fold difference between retardation factors between bedrock and deposit is a dominating factor for the distribution of particle travel times. It could be justified by a reference to other workgroups who found similar retardation factor ratios between rock and deposit.
147: Extrapolating a decay coefficient from measurements at 3 and 7cm to a depth of 5m is questionable. The assumption that the minimum conductivity is 10^-6 (m/s) probably determines the decay coefficient much more than the actual measurements. The only reference is, again, only a single study from the own workgroup. The authors should be able to find more measurements of sediment conductivities in the literature to strengthen their assumption.
159 – 173: In line 165 you state that the realistic boundary condition would be a recharge-controlled boundary for most of the terrain. However, you choose a head boundary condition instead and use a mesh-coarsening algorithm to fit the recharge. Why didn’t you simply use a recharge-controlled boundary? Coarsening the mesh to fit a result is somewhat unorthodox. All discretizing simulation techniques have in common that an infinitely fine mesh resolution results in the exact solution of the underlying differential equations. The boundary condition in this study, however, implies a tradeoff between model inaccuracy and boundary condition inaccuracy, which, in my opinion, should be avoided by choosing mesh-size independent boundary conditions.
177: A figure of the mesh/domain would be helpful. The domain is rectangular? I originally thought it was a whole (sub-)catchment with its natural boarders (and lateral no-flow boundary conditions).
179: Were the particles weighted somehow? In the following particle statistics, what does one particle stand for? A certain fraction of groundwater volume? A certain area/volume of bedrock? Please indicate why your choice is the best choice for the research question.
196: is “c” the same coefficient as “delta” in line 147?
213: Why did you use local regions for downscaling? Couldn’t it also be a 100x100m region from somewhere else in the world? Or do you assume that there is a correlation between the catchment topography and its streambed-topograpy? I don’t think that Wörman, et al., 2007a proved a local correlation between topographies. I think it should be clarified for the reader if a local correlation is assumed or if the regional topography is simply used as a sophisticated random field generator and the topography data could also be taken from somewhere else.
179/246: Both particle tracings should be described in a single chapter.
248: If intermediate particle traces are those that did not enter the bedrock and deep particles are those that started at 500m depth, you miss the flow that enters the bedrock but not to a depth of 500m. Superimposing two particle tracings with different seeds and without weighting them properly against each other adds some randomness to the results.
260/Fig2/179: The release plane for particles was described as “approximately 500 m below the minimum topography elevation”. That describes a flat plane. However, the release surface in Fig.2 two shows a curvy plane. Which one is correct and why is the depth “approximately”?
Fig 3 and Fig 6 and the corresponding text could also be placed in the methods section.
Fig 7 and corresponding text: How many particle traces from the deep fraction entered the 5x5m domain to be superimposed by hyporheic flux? According to Fig4 it could only be a handful. Why are the dashed lines so smooth and the solid lines are not? Do they represent the same amount of particle traces?
377-400: To be honest, I did not understand what you did here. What is on the y-axis of the CDF? The whole topic “coherent area” needs a better explanation. Why did you choose coherence as a measure? What environmental process would coherence be important for? My interpretation is that you created a list of coherent (however coherence was defined precisely) upwelling patches and calculated their surface area. Now you found, for example, that 50% of these patches had surface areas > 400m². Is that correct? If so, I would strongly recommend not to use a CDF for presentation but a PDF. Nothing is cumulating here, CDFs are easily interpreted as “number of particle traces that reached the surface” where 100 means all particles exited the domain or something similar.
409: Why does the scenario assume accumulation? Shouldn’t it be steady state in- and outflow at some point?
415: I’m not familiar with dose assessment but if it is based on the idea that groundwater upwelling happens on a large area without fragmentation it is obviously oversimplifying groundwater flow. If so, you should explain the dose model in more detail and propose an improvement to the model.
Boano, F., Revelli, R., and Ridolfi, L. (2008), Reduction of the hyporheic zone volume due to the stream-aquifer interaction, Geophys. Res. Lett., 35, L09401, doi:10.1029/2008GL033554.
Trauth, N., Schmidt, C., Vieweg, M., Maier, U., and Fleckenstein, J. H. (2014), Hyporheic transport and biogeochemical reactions in pool-riffle systems under varying ambient groundwater flow conditions, J. Geophys. Res. Biogeosci., 119, 910– 928, doi:10.1002/2013JG002586.
Fox, A., Laube, G., Schmidt, C., Fleckenstein, J. H., and Arnon, S. (2016), The effect of losing and gaining flow conditions on hyporheic exchange in heterogeneous streambeds, Water Resour. Res., 52, 7460– 7477, doi:10.1002/2016WR018677.
Bhaskar, A. S., Harvey, J. W., and Henry, E. J. (2012), Resolving hyporheic and groundwater components of streambed water flux using heat as a tracer, Water Resour. Res., 48, W08524, doi:10.1029/2011WR011784.
Citation: https://doi.org/10.5194/hess-2021-148-RC3 -
AC5: 'Response to the comments of Referee #3', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC5-supplement.pdf
-
AC5: 'Response to the comments of Referee #3', Brian Babak Mojarrad, 10 Jul 2021
Status: closed
-
CC1: 'Comment on hess-2021-148', Ulrik Kautsky, 17 May 2021
The paper presents an interesting model study concerning potential effects of the hyporheic zone on groundwater flow patterns using a selection of the hydrological data available for Krycklan. In its present state the study is theoretical. The study would be strengthened if it included an analysis describing the extent to which the regional model was able to capture the hydrology at Krycklan. This would help to address questions about the validity of the regional hydrological model and may help frame results from the hyporheic, local-scale simulations as applicable to a real-world scenario.
The paper would be further strengthened if it discussed the robustness of the results as affected by the parameterizations and structural assumptions made in the numerical and conceptual models. For example, the potential effects of landscape topography, time-variation of streambed topography, rock fractures, soil stratification, spatial heterogeneity of streambed sediments, parameterization of the infiltration rate at the surface, and model boundary conditions are not discussed. Without such a discussion it cannot be deduced the extent to which variabilities in these, and other, variables will affect the hyporheic phenomena postulated in the modelling exercise; it is therefore not possible to assess the extent to which the results and/or methodologies presented in the study are relevant to other sites.
Citation: https://doi.org/10.5194/hess-2021-148-CC1 -
RC2: 'Review of article', Anonymous Referee #2, 29 May 2021
Review Hess hyporheic fluxes
General discussion of the content of the paper “The influence of hyporheic fluxes on regional groundwater discharge zones”
The paper "The influence of hyporheic fluxes on regional groundwater discharge zones" submitted to HESS by Mojarrad et al. investigates different scale flow systems by means of topographically induced groundwater circulation and the role of hyporheic flow on groundwater circulation in the river valleys. The flow systems are discussed in relation to possible accumulation of radioactive waste originating from deeper aquifer regions within the Quaternary deposits; by differentiating between re- and discharge zones.
I think that the approach, of the influence of topography on the distribution of hydraulic potential in different scales is interesting. The extent to which the link of deep groundwater circulation with the potential field of the hyporheic interstitial can be easily connected would have to be discussed in more depth, especially by delving into the processes that can also influence the potential fields at different scales.
The concept underlying the work is not new and it is surprising that the authors do not mention the fundamental work of J. Toth.
The authors cite the work of W. Zijl, who, through a Fourier analysis and consideration of the anisotropy of the geologic sequences, or important boundaries, such as the topography of the bedrock surface, or different structural properties of the bedrock, determine a series of flow cells. A reduction to two or three as proposed in this paper is probably too simplified. Thus, I would find it justified for the authors to emphasize the conceptual aspect of the work and to discuss the role of influence of topography of the river bed (i.e., the river corridor concept of Stanford and Ward), the influence of heterogeneity of the geological sequence, the role of the relative water flux contributions etc.).
Based on the work (concept) of Toth, it is not surprising that the drainage system can correspond to the exfiltration zones.
However, there are other aspects to be considered:
The River corridor concept of Stanford and Ward, shows that the topography of the riverbed and the topography of an impounding layer (can be e.g. glacial deposits, bedrock surface or discontinuities of the gradient of the riverbed along a river course) produce infiltration and exfiltration zones within the hyporheic interstitial, which can be very dominant with respect to water fluxes.
I do not know the glacial sediments of the area, but I suspect that because of the diversity of processes, heterogeneities in these deposits also lead to vertical hydraulic gradients that could significantly affect the simple potential distribution.
The structural heterogeneity of the bedrock and the character of the hydraulically relevant structures of the subsurface (shear zones, fracture patterns, etc.) can also significantly affect the anisotropy of the hydraulic conductivities, so that over long geologic times the pattern of the exfiltration zones is also not necessarily uniformly distributed along the drainage system.
The concept depth decaying hydraulic conductivity has to be approved by regional specific data. Other, more complex heterogeneity development with depth could have a strong influence on the Potential distribution.
A much more important point influencing the regional flow systems over longer times periods (transport of radionuclides from certain depths of the bedrock) is the dynamics of the development of the topography. Please discuss, how topography was shaped during the last 15’000 years. Over long time periods, topography cannot be assumed to constant.
I think the paper could benefit from a more in-depth discussion. In its current form, the argumentation is a bit too simplistic.
Some general comments:
- Although the character of the model is mostly conceptual the authors state that results from intense field investigations exist. Nevertheless, nearly none of these existing data are specified or used for calibration and/or validation of the model (i.e. character of the hyporheic zone, heterogeneity character of the glacial deposits.
-For seven different soil types in the Krycklan catchment hydraulic conductivity was obtained. A sensitivity analysis for hydraulic conductivity of streambed sediment would be interesting.
- Provide a more quantitative of visualization of discharge locations (Fig. 1), e.g. by means of point densities. Likewise, an illustration with the "pinholes" of groundwater discharge and/or "nested" flow cells (maybe for a zoom) would also help to understand the different flow processes.
- How was the catchment area delineated (surface?). And is this approach appropriate when evaluating the deep aquifer, when the shape of the topography changed over the last 15’000 years? A 3D visualization would help to better understand the geological settings in relation to the topography.
- Some repetitions could be avoided, like e.g. Software use.
Some specific comments:
- 29: Definition of "long-term"
- 54: This strongly depends on the geology which means that without a regional geological model a detailed statement due to deep groundwater discharge zones remain fragmentary
- 66: please add References of S. Todd
- 84: the history of the hydrology at different time scales will also influence the contribution of water from different flow systems in the exfiltration zones.
- 82: I cannot find information on geological heterogeneity
- 86: Darcy's law should be known to the readership?
- 103: Missing specific information on the time scale and the rates of considered processes
- 118: There should be drillings for geological information’s. A corresponding map and a stratigraphic-lithologic overview, allowing to evaluate the degree of vertical variability of hydraulic properties is missing.
- 122: a sedimentological decription of the glacial sediments would allow a better info on heterogeneity, till compris a lot of different glacial sediment types
- 134: There is no information about the geometric extension of the model and why was Comsol chosen for such a simple model of the geological setting and not modflow, feflow or any groundwater affine software?
- 137: The presented material does not allow this simplification, which project data support the statements of the cited publications?
ions
l.158: The information of the mesh sizes should be expanded with the number and delaunay quality of elements
- 175: DT has not been introduced
- 178: give a reference to the particle tracing routine used
l 181: One of the described aims in the introduction is the influence of radionuclide spreading for humans and biota. With a residence time of 320 million years it would be hard to make any predictions at all? A cumulative curve from particles reaching the surface over time could be more detailed.
At the long time scale the topography has changed and as a result the flow field also would have changed,
- 219-220: Specify "other models and statistical uncertainties"
- 239: Correct "Monte Carla"
- 335: Can you better explain the role of the Froude number on the dynamic head component?
- 348-350: Why you mention this? Is it not expected?
- 495: In the long time, changes of the Surface and streambed morphology is expected, therefore it is not clear how these changes interfere with the flow field in the deeper deposits (Quaternary and bedrock).
L 510 and following lines: Die you take into account structural aspects of the heterogeneity in the bedrock, such as heterogeneity due to shear zones , mylonites etc.
Figures:
Integrate Figures 1 & 2
4: Up-date: Northing? Points instead of stars?
Figure 9: Correct catchemnt
Citation: https://doi.org/10.5194/hess-2021-148-RC2 -
AC2: 'Response to the comments of Referee #2', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC2-supplement.pdf
-
AC1: 'Response to the Community Comments #1', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC1-supplement.pdf
-
RC2: 'Review of article', Anonymous Referee #2, 29 May 2021
-
RC1: 'Comment on hess-2021-148', Anonymous Referee #1, 20 May 2021
General comments
This paper investigates how hyporheic flow redistributes regional discharge at the scale of riverbeds in a case study and discusses the implications for the fate and transport of radionuclides from deep nuclear waste depositories. The study involves a multiscale flow and transport modelling framework and a rather sophisticated analysis of model outputs.
The main result (fragmentation of the regional discharge by hyporheic flow) is rather intuitive and could already be anticipated by looking at Tóth (1963) flow fields. This takes nothing away from the merit of the study, which achieved a proper demonstration and quantification of this phenomenon in a realistic example. Therefore, I recommend publication without any doubt.
The main issue I found is that the paper lacks a number of explanations (see below), but I am sure the authors can improve on this aspect.
Detailed comments
L51-56: I am not sure that velocity is the most relevant indicator for what you are trying to convey here. In fact, all the flow paths end up having similar velocities when approaching discharge (Cardenas and Jiang, 2010; Zijl, 1999; Zlotnik et al., 2011). Instead, I would think that the ratio of hyporheic flow (i.e., its total flow rate) to that of regional flow is more relevant in this discussion.
L59: I guess you mean “principal effects” and not “principle effects”.
L62-64: The sentence is grammatically incorrect.
L110: This section should be better put before the description of the models.
L111-114: A situation map would be useful (I would suggest including Figure S1 here).
L129: This sentence is unclear. I guess you mean: the mean annual runoff estimated from the stream discharge measurements was set as the infiltration (please correct if needed).
L129: Is it reasonable to neglect overland flow in your study area (you may be overestimating infiltration)?
L129: Can you indicate the calculated infiltration rate?
L161-162: I suggest referring to Haitjema and Mitchell-Bruker (2005) in support of this sentence.
L162-164: I suggest referring to Bresciani et al. (2016a, 2016b) in support of this sentence.
L166-173: This only makes sense if hydraulic head is specified and equal to the topography along the top boundary, but you just said above that you are using a recharge condition, so I am lost here.
L134-185: What are the horizontal limits of the domain?
L194: I do not understand the meaning of “Cdamp(λi)”. Is Cdamp a function of λi (I would think not since Cdamp seems to be treated as a constant)? And if it is, shouldn’t it be λij?...
L206-216: How does this relate to the previous paragraph?
L187-216: What are the extent and boundary conditions of the hyporheic flow model? I guess the boundary conditions must be head = 0 everywhere but the top boundary so as to keep a continuous solution when doing the superposition...?
L219: The term “models” is confusing here. I guess you refer to the other parameters of the hyporheic flow model and all the parameters of the catchment-scale model.
L239: “Carlo”, not “Carla”.
L247: What are “the” cubes? You have not talked about cubes before.
L252: What does “the 1552” refer to?
L253-256: How did you distinguish between intermediate and deep groundwater flow paths from these particles (did you track them backward as well)? Furthermore, how can you be sure that some of these particles are not hyporheic flow?
L264-272: This part could be clearer. Did you focus on the same 1552 areas as above (I guess so)? How did you determine the coherent catchment-scale discharge areas (I guess this would involve particle tracking and a certain grouping method)?
L308-310: I think the differences between the three layers are mostly independent of the hierarchical structure of flow cells (which was not evaluated in this study, by the way).
L331: Define the Froude number.
L418-420: So is it a good news (less exposure time of aquatic sediments)?
L421: Why would it lead to higher exposure if the exposure time is shorter?
References
Bresciani, E., Gleeson, T., Goderniaux, P., de Dreuzy, J.R., Werner, A.D., Wörman, A., Zijl, W., Batelaan, O., 2016a. Groundwater flow systems theory: Research challenges beyond the specified-head top boundary condition. Hydrogeol. J. 24, 1087–1090. https://doi.org/10.1007/s10040-016-1397-8
Bresciani, E., Goderniaux, P., Batelaan, O., 2016b. Hydrogeological controls of water table-land surface interactions. Geophys. Res. Lett. 43, 9653–9661. https://doi.org/10.1002/2016gl070618
Cardenas, M.B., Jiang, X.-W., 2010. Groundwater flow, transport, and residence times through topography-driven basins with exponentially decreasing permeability and porosity. Water Resour. Res. 46, W11538. https://doi.org/10.1029/2010wr009370
Haitjema, H.M., Mitchell-Bruker, S., 2005. Are water tables a subdued replica of the topography? Ground Water 43, 781–786. https://doi.org/10.1111/j.1745-6584.2005.00090.x
Tóth, J., 1963. A theoretical analysis of groundwater flow in small drainage basins. J. Geophys. Res. 68, 4795–4812. https://doi.org/10.1029/JZ068i016p04795
Zijl, W., 1999. Scale aspects of groundwater flow and transport systems. Hydrogeol. J. 7, 139–150. https://doi.org/10.1007/s100400050185
Zlotnik, V.A., Cardenas, M.B., Toundykov, D., 2011. Effects of multiscale anisotropy on basin and hyporheic groundwater flow. Ground Water 49, 576–583. https://doi.org/10.1111/j.1745-6584.2010.00775.x
Citation: https://doi.org/10.5194/hess-2021-148-RC1 -
AC3: 'Response to the comments of Referee #1', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC3-supplement.pdf
-
AC3: 'Response to the comments of Referee #1', Brian Babak Mojarrad, 10 Jul 2021
-
CC2: 'Comment on hess-2021-148', Ulrik Kautsky, 31 May 2021
We had an additional important comment to the manuscript lost in my previous post
Many of the claims regarding ecology and human health are outside of the scope of the study especially given the simplifications that have been made regarding the hydrology of the biosphere. An attempt to extrapolate results into these areas is not properly motivated. Results of the study will stand for themselves without having to extrapolate them into areas outside of the study’s scope which, at present, is largely theoretical.
Citation: https://doi.org/10.5194/hess-2021-148-CC2 -
AC4: 'Response to the Community Comments #2', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC4-supplement.pdf
-
AC4: 'Response to the Community Comments #2', Brian Babak Mojarrad, 10 Jul 2021
-
RC3: 'Comment on hess-2021-148', Anonymous Referee #3, 31 May 2021
General
The authors of this study used a complex multiscale model to compare the travel time and spatial distribution of catchment scale groundwater flow with and without the influence of hyporheic flow. The results are loosely linked to the fate of nuclear particles that might potentially leak out from nuclear waste disposal in the bedrock of the catchment. The main finding of the study is the fragmentation of the groundwater discharge areas, the reduction of the effective volume that carries groundwater flow and the consequential reduction of the groundwater travel time in the upper 5m of the domain. A monte carlo approach was used to some extend to generate a distribution of possible outcomes where exact model parameters were unknown.
The underlying model is without a question quite sophisticated. However, I wonder if the right model was chosen to answer the specific scientific question. The main finding, that the discharge zone is fragmented, is relatively obvious and has been published various times in model studies from the hyporheic perspective (e.g. Boano et al 2008, Trauth et al 2014, Fox et al 2016) and also investigated experimentally (e.g. Bhaskar 2012). Given that the general effect is well known, I would have expected an in-depth analysis of the correlations of at least some of the parameters involved. I wonder if this complex, multiscale, pseudo-realistic model is the right choice to draw systematic conclusions with results that have a direct use for other scientists. The model has too many unknowns to investigate general findings, e.g. correlating discharge area reduction to the ratio of hyporheic head and groundwater head or something similar.
If the authors didn’t aim to draw general results but rather calculate the effect for a specific case (nuclear waste disposal in this specific catchment), I see two problems. First, case studies do not match the scope of the HESS journal. Second, some of the boundary conditions have been selected without verification or are too unclear to be used for a specific use case (especially a safety relevant topic like nuclear waste disposal). Generally, I think that a lot of choices for boundary conditions are justified only by references to past work of the own workgroup. In some cases, references to independent workgroups would strengthen the trustworthiness of the model (see detailed comments below).
The second major finding is the reduced travel time within the upper 5m of the 500m deep domain. Also here, I miss a systematic investigation of the underlying mechanism. E.g. what is the correlation between the fragmentation and the travel time? What is the correlation with the hyporheic head and groundwater head? Is the conductivity a relevant factor in this correlation? These kind of questions should be answered with specific results rather than vague discussions, because the vague answers to these questions are obvious. Again, it would have been much easier to answer these kind of questions if the model was less complex.
In addition to the missing universal results, I don’t understand why the reduction of travel time in the upper 5m is deemed relevant. The effect of the hyporheic zone decreases exponentially with depth (see eq(5) or Elliot & Brooks 1997a), which is why only the very last segment of a particles travel path will be influenced. In Line 353 it says that the effect of the hyporheic head is strongest, where the overall seepage velocity is low. That means that particles that traveled centuries to millennia (Fig 5) to reach the surface will lose a few years on their last few meters (Fig 7). Even when the different retention coefficients in bedrock and deposits were considered (missing in Fig 7), the effect should be minor. I miss a clear explanation what processes are potentially influenced by the change in travel time in this last stage of the streampath.
Overall, I’m afraid that the study design is not able to answer the research questions to a degree that goes beyond the intuitive and well-published findings.
Specific:
102: I don’t think it is necessary to show the definition of Darcy’s velocity, but if you do, please use the already defined symbol for Darcy’s velocity “W_c” instead of q and q_seepage.
108: The 50-fold difference between retardation factors between bedrock and deposit is a dominating factor for the distribution of particle travel times. It could be justified by a reference to other workgroups who found similar retardation factor ratios between rock and deposit.
147: Extrapolating a decay coefficient from measurements at 3 and 7cm to a depth of 5m is questionable. The assumption that the minimum conductivity is 10^-6 (m/s) probably determines the decay coefficient much more than the actual measurements. The only reference is, again, only a single study from the own workgroup. The authors should be able to find more measurements of sediment conductivities in the literature to strengthen their assumption.
159 – 173: In line 165 you state that the realistic boundary condition would be a recharge-controlled boundary for most of the terrain. However, you choose a head boundary condition instead and use a mesh-coarsening algorithm to fit the recharge. Why didn’t you simply use a recharge-controlled boundary? Coarsening the mesh to fit a result is somewhat unorthodox. All discretizing simulation techniques have in common that an infinitely fine mesh resolution results in the exact solution of the underlying differential equations. The boundary condition in this study, however, implies a tradeoff between model inaccuracy and boundary condition inaccuracy, which, in my opinion, should be avoided by choosing mesh-size independent boundary conditions.
177: A figure of the mesh/domain would be helpful. The domain is rectangular? I originally thought it was a whole (sub-)catchment with its natural boarders (and lateral no-flow boundary conditions).
179: Were the particles weighted somehow? In the following particle statistics, what does one particle stand for? A certain fraction of groundwater volume? A certain area/volume of bedrock? Please indicate why your choice is the best choice for the research question.
196: is “c” the same coefficient as “delta” in line 147?
213: Why did you use local regions for downscaling? Couldn’t it also be a 100x100m region from somewhere else in the world? Or do you assume that there is a correlation between the catchment topography and its streambed-topograpy? I don’t think that Wörman, et al., 2007a proved a local correlation between topographies. I think it should be clarified for the reader if a local correlation is assumed or if the regional topography is simply used as a sophisticated random field generator and the topography data could also be taken from somewhere else.
179/246: Both particle tracings should be described in a single chapter.
248: If intermediate particle traces are those that did not enter the bedrock and deep particles are those that started at 500m depth, you miss the flow that enters the bedrock but not to a depth of 500m. Superimposing two particle tracings with different seeds and without weighting them properly against each other adds some randomness to the results.
260/Fig2/179: The release plane for particles was described as “approximately 500 m below the minimum topography elevation”. That describes a flat plane. However, the release surface in Fig.2 two shows a curvy plane. Which one is correct and why is the depth “approximately”?
Fig 3 and Fig 6 and the corresponding text could also be placed in the methods section.
Fig 7 and corresponding text: How many particle traces from the deep fraction entered the 5x5m domain to be superimposed by hyporheic flux? According to Fig4 it could only be a handful. Why are the dashed lines so smooth and the solid lines are not? Do they represent the same amount of particle traces?
377-400: To be honest, I did not understand what you did here. What is on the y-axis of the CDF? The whole topic “coherent area” needs a better explanation. Why did you choose coherence as a measure? What environmental process would coherence be important for? My interpretation is that you created a list of coherent (however coherence was defined precisely) upwelling patches and calculated their surface area. Now you found, for example, that 50% of these patches had surface areas > 400m². Is that correct? If so, I would strongly recommend not to use a CDF for presentation but a PDF. Nothing is cumulating here, CDFs are easily interpreted as “number of particle traces that reached the surface” where 100 means all particles exited the domain or something similar.
409: Why does the scenario assume accumulation? Shouldn’t it be steady state in- and outflow at some point?
415: I’m not familiar with dose assessment but if it is based on the idea that groundwater upwelling happens on a large area without fragmentation it is obviously oversimplifying groundwater flow. If so, you should explain the dose model in more detail and propose an improvement to the model.
Boano, F., Revelli, R., and Ridolfi, L. (2008), Reduction of the hyporheic zone volume due to the stream-aquifer interaction, Geophys. Res. Lett., 35, L09401, doi:10.1029/2008GL033554.
Trauth, N., Schmidt, C., Vieweg, M., Maier, U., and Fleckenstein, J. H. (2014), Hyporheic transport and biogeochemical reactions in pool-riffle systems under varying ambient groundwater flow conditions, J. Geophys. Res. Biogeosci., 119, 910– 928, doi:10.1002/2013JG002586.
Fox, A., Laube, G., Schmidt, C., Fleckenstein, J. H., and Arnon, S. (2016), The effect of losing and gaining flow conditions on hyporheic exchange in heterogeneous streambeds, Water Resour. Res., 52, 7460– 7477, doi:10.1002/2016WR018677.
Bhaskar, A. S., Harvey, J. W., and Henry, E. J. (2012), Resolving hyporheic and groundwater components of streambed water flux using heat as a tracer, Water Resour. Res., 48, W08524, doi:10.1029/2011WR011784.
Citation: https://doi.org/10.5194/hess-2021-148-RC3 -
AC5: 'Response to the comments of Referee #3', Brian Babak Mojarrad, 10 Jul 2021
The comment was uploaded in the form of a supplement: https://hess.copernicus.org/preprints/hess-2021-148/hess-2021-148-AC5-supplement.pdf
-
AC5: 'Response to the comments of Referee #3', Brian Babak Mojarrad, 10 Jul 2021
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 1,301 | 607 | 69 | 1,977 | 151 | 53 | 60 |
- HTML: 1,301
- PDF: 607
- XML: 69
- Total: 1,977
- Supplement: 151
- BibTeX: 53
- EndNote: 60
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 687 | 36 |
| China | 2 | 387 | 20 |
| Germany | 3 | 162 | 8 |
| Sweden | 4 | 130 | 6 |
| France | 5 | 61 | 3 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 687