the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Jan 2024
Improving the internal hydrological consistency of a process-based solute-transport model by simultaneous calibration of streamflow and stream concentrations
Abstract. Improving the consistency of hydrological models, i.e. their ability to reproduce observed system dynamics, is required to increase their predictive power. As the use of streamflow data for calibration is necessary but not sufficient to constrain model and warrant model consistency, other strategies must be considered, in particular the use of additional data sources. The aim of this study is to test whether simultaneous calibration of dissolved organic carbon (DOC) and nitrate (NO3-) concentrations along with streamflow improves the hydrological consistency of a parsimonious solute-transport model. A multi-objective and multi-variable approach was used to evaluate the model in an intensive agricultural headwater catchment. Our results showed that using daily stream concentrations of DOC and NO3- together with streamflow data during calibration did not improve the model's ability to accurately predict streamflow for calibration or evaluation periods. However, the internal consistency of the model was improved for the simulation of low flows, groundwater storage and upstream soil storage, but not for the simulation of riparian soil storage. Parameter uncertainty decreased when the model was calibrated using solute concentrations, except for parameters related to fast and slow reservoir flow. This study shows the added value of using multiple data sources in addition to streamflow data for calibration, in particular DOC and NO3- concentrations, to constrain hydrological models for a better representation of internal hydrological states and flow. With the increasing availability of solute data from catchment monitoring, this approach provides an objective way to improve the internal consistency of hydrological models that can be used with confidence in scenario evaluation.
- Preprint
(9776 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on hess-2023-292', Anonymous Referee #1, 19 Feb 2024
Review on Improving the internal hydrological consistency of a process-based solute-transport model by simultaneous calibration of streamflow and stream concentrations
In their manuscript, Salmon-Monviola et al. explore the utilization of dissolved organic carbon (DOC) and nitrate concentrations as constraints to refine streamflow predictions and enhance the internal consistency of a conceptual hydrological model. While their investigation revealed that DOC and nitrate concentrations did not enhance streamflow predictions per se, they did, however, reduce the uncertainty in model parameters and the representation of internal hydrological states and flow. The manuscript is notably well-written and clearly makes a case for improving the internal consistency of the conceptual hydrological models by adding additional constraints, such as solute concentrations. It convincingly illustrates how the inclusion of DOC and nitrate concentrations considerably affects the representation of underlying hydrological states and flow. Nonetheless, I have some doubts about whether the constrained version reflects greater realism, as it improved the representation of groundwater storage but showed equal to even worse results for soil moisture (more details in the general comments below). With these concerns addressed, the manuscript holds promise as a significant contribution to the readership of HESS.
General comments:
I find it convincing that the simulations, constrained by nitrate concentrations (S3 and S4), have improved the representation of groundwater levels. However, I cannot entirely follow the interpretation of an improved representation of upslope soil moisture. The authors reference Figure 11 to support this claim, yet upon examination, I observe only marginal disparities between the non-constraint simulation (S1) and the one constrained by DOC (S2), while S3 and S4 exhibit notably lower performance in terms of NSE, KGE, PBIAS, and RSME. Consequently, from my perspective, none of the simulations incorporating DOC and/or nitrate concentrations (S2 – S4) consistently elevate internal model consistency in representing soil moisture and groundwater. Hence, it remains uncertain whether these simulations merely exert a general influence on the representation of hydrological states and flows or, indeed, foster an overarching enhancement in the models' internal consistency. It might be that the representation of DOC and nitrate processing and transport are too simple, or soil moisture measurements are not presentative for the entire catchment, as nicely discussed in chapters 4.2 and 4.3. Nevertheless, I find this point insufficiently addressed.
I missed a discussion on the applicability across different catchments, especially in the view that dominant sources and pathways of DOC and nitrate concentrations can strongly vary in other settings. Based on my interpretation of Figures 5 and 6, it seems that DOC concentrations exhibit an enrichment pattern (i.e., increasing concentrations with increasing streamflow), whereas nitrate concentrations demonstrate dilution patterns (i.e., decreasing concentration with increasing discharge). However, these patterns may differ significantly in other catchments with distinct sources and pathways (e.g., Winter et al., 2021; Knapp et al., 2020). How might these differences affect your model setup? From my perspective, it is crucial to address the implications of the specific catchment characteristics and the transferability of your findings to other catchments.
Specific comments:
Line 95 - 96: What exactly do you mean by substantially differing? Is there a directional relationship between concentrations and discharge, or are dynamics completely independent?
Line 98: I recommend toning down a little to “can” be closely related.
Line 98 – 104: See here my second general comment. There are various patterns of DOC nitrate export dynamics depending on the storage and flow paths within the catchment. This is a little too simplified for my taste.
Line 105 – 108: Nice and clear!
Line: 116: Could you write the full name of AgrHyS once, please?
Line 135: Why only in riparian-zone soils and not in all soils? What dynamic do you infer if speaking of unlimited supply? Chemostasis? Enrichment? Information about DOC sources and the relationship between Q and DOC, and Q and nitrate (for example in the SI) could help to back up this argument.
Line 379: Those deep infiltration losses appear a little bit like a “mathematical marionette” to me, and they appear to be a highly sensitive parameter. Can you please elaborate on those a little more?
Line 454 – 457: This sentence is very hard to read. Can you rewrite or slit it, please?
Line 527 - 529: See my first major comment. The improvement in groundwater storage is convincing, but I do not see a significant improvement in upslope soil moisture. Here, S1 performs similarly well to S2 and clearly better than S3 and S4. I find it critical that S1 seems to perform better for soil moisture, while S3 and S4 perform better for groundwater. Thus, neither DOC nor NO3 seems to consistently improve the internal representation of water fluxes.
Line 541 – 543: Consequently, this appears overstated to me. I only see that it improves the groundwater storage representation, while two out of three scenarios show a lower performance for soil moisture.
Line 556 – 575: I agree with all points discussed here, but I am not entirely convinced that adding DOC and NO3 made the model produce the ‘right answers for the right reason’ for the reasons mentioned above.
602 – 604: Good point. I also enjoyed reading chapter 4.2 and 4.3.
Line 667 - 672: This appears contradictory to me. Did it or did it not improve the models' ability to reproduce streamflow? The first sentence says no, the second yes.
Figures
Figure 1: Why show the entire Narzin catchment if your results focus on the Kervidy-Narzin catchment only? In the caption, you do not mention the Narzin catchment either – It seems more straightforward to me to show the Kervidy-Narzin catchment only.
Figures 4, 5, and 6: Adding the scenario names (S1 – S4) to the figures would improve clarity compared to only mentioning them in the description below. Moreover, I find it difficult to see the differences between the model runs. You might consider adding an observed vs. simulated plot with a 1:1 line to the SI. But this is just a suggestion.
Figure 9: consider using a white font in front of the dark blue box for S1. A black font is hard to read with dark background (this also applies to the other figures).
Figures 9, 10, and 11 could be combined and reduced in height to save space.
Overall, I do think 12 Figures, each of them taking up around half a page, is too much. This might be a matter of taste, but I would recommend combining figures and lowering their height and/or shifting some of them to the SI.
References
Knapp, J. L., Freyberg, J. von, Studer, B., Kiewiet, L., and Kirchner, J. W.: Concentration-discharge relationships vary among hydrological events, reflecting differences in event characteristics, Hydrol. Earth Syst. Sci. Discuss., 1–27, https://doi.org/10.5194/hess-24-2561-2020, 2020.
Winter, C., Lutz, S. R., Musolff, A., Kumar, R., Weber, M., and Fleckenstein, J. H.: Disentangling the impact of catchment heterogeneity on nitrate export dynamics from event to long‐term time scales, Water Resour. Res., 57, e2020WR027992, https://doi.org/10.1029/2020WR027992, 2021.
Citation: https://doi.org/10.5194/hess-2023-292-RC1 - AC1: 'Reply on RC1', salmon jordy, 12 Apr 2024
-
RC2: 'Comment on hess-2023-292', Anonymous Referee #2, 06 Mar 2024
The study by Salmon-Monviola et al. investigates the utility of using biogeochemistry parameters to improve a conceptual hydrological model's accuracy and internal consistency. The authors modeled hydro-biogeochemical processes in the Kervidy-Naizin catchment in NW France to test this idea. They used a hydrological model based on previously published models, e.g., Hrachowitz et al. (2015). Adding DOC and NO3- processes into the model did not improve streamflow prediction. However, adding solute processes to streamflow improved the model's internal consistency, as demonstrated by the ability to model groundwater level and upslope soil moisture and reduce parameter uncertainty. The paper was generally well organized and well written, and the aims of the paper are within the scope of HESS. The paper requires some revisions, but if done satisfactorily, this paper would be a solid addition to the literature.
General comments:
I agree with Reviewer 1's ‘General Comments’ section. Figure 11 does not provide strong enough evidence to suggest that adding DOC to the model (S2) improves model (S1) accuracy. I would also like to echo the need for discussion on the applicability of these findings across regions.I also found it difficult to follow how results were produced and interpreted. This may be because I don’t have experience with this type of modeling. However, HESS has a broad hydrological readership, so the following comments are aimed to help communicate to a broader audience. For example, it wasn't easy to follow how the DOC and NO3 parameters were incorporated into the conceptual model. The hydrology was well outlined in Figure 2, and it would help to have a similar figure (perhaps Fig. 2b) for the biogeochemistry. Secondly, many model performance metrics (e.g., PBIAS, RMSE, eCDF, etc.) were not adequately explained or introduced, making it challenging to follow model interpretations. Additionally, the results alluded to some statistics (e.g., line 515), but no statistical methods were described. Another inconsistency was in the figures. Figures 4-6 had distinct calibration and evaluation periods for streamflow, DOC, and NO3 concentrations. Yet for figures 9-11 showing groundwater level and soil moisture, the whole period appeared to be for evaluation, but the date ranges on the x-axis varied. It wasn’t clear how these modeling approaches differed and why the date ranges changed.
Finally, as a stylistic point, I found all the model abbreviations and acronyms in the text highly distracting. I suggest that to improve readability, the authors should only use common abbreviations (e.g., DOC) in the text.
Specific comments:
Lines 87-88: vague sentence. The example needs to be more concrete.
Lines 95-97: What is “This potential” referring to? The spatial distribution of solutes where? In groundwater?
Line 130: are ‘livestock units’ different from the number of animals?
Line 157: define TDR
Lines 219-223: need more clarification in this section. Are you suggesting more N is removed in winter via denitrification, and in summer by biological uptake?
Lines 232-233: The two Birkel et al. papers cited here are from catchments with wetlands supplying the bulk of the DOC. So, it isn’t surprising that the stream DOC reactivity is negligible. However, DOC is typically more reactive in agricultural catchments (see (Shang et al 2018, Eder et al 2022)). However, it doesn’t seem like stream water is a reservoir with an associated transit time in the conceptual model, so how would stream water DOC (and NO3) reactivity be included? It seems like your model is well-positioned to speak to solute groundwater transport, but less so for surface water.
Lines 235-236: I think this is a fair assumption, but I would reword the justification to say that deeper mineral soils are DOC sinks.
Line 240: why is the ‘L’ in [M L-3] cubed? What does the L stand for? It seems to represent units of volume, but it wouldn’t make sense for that to be cubed. Please clarify and be consistent throughout the text.
Lines 281-288: two very long sentences. Consider revising it into multiple sentences.
Line 322: start new paragraph.
Line 324: please define which 6 metrics were used.
Line 345: define ‘Pareto front’ and how to interpret.
Line 393: the phrase “relatively well” is vague. Be more specific.
Lines 401-403: Consider reporting the mean and standard deviation of observed DOC and NO3 concentrations to aid RMSE interpretation. In some catchments, mean DOC and NO3 concentrations are less than or equal to your observed RMSE, so it’s important to contextualize this information.
Lines 459-461: Great sentence!
Line 478: move your key result to the beginning of this paragraph.
Line 490: move your key result to the beginning of this paragraph.
Line 491: confusing use of ‘well’—is it an adverb or noun here?
Table 2: In the DOC concentration row, Definition column, remove the word ‘rate’ in “DOC concentration rate…”
Figure 4: the way this is plotted, it’s difficult to tell when the observed and simulated data are just the same or there are gaps in the data.
Figure 6: when the model predicts highest annual concentrations there is no observational data. What happened to that data? Is it missing? This needs to be explained in the methods, and some discussion is needed to explain how this might affect model accuracy.
Figure 7: the last sentence of the caption states that “The boxplot of KGENO3 for scenarios S1 and S2 are absent because their values were negative.” What do negative values mean for model performance?
Figure 8: text is too small in the graphs. It might also be helpful to explain how to interpret the eCDF values somewhere in the manuscript.
Figures 9-11: The background color in the table insert makes the text illegible. Also, font size is too small.References
Eder A, Weigelhofer G, Pucher M, Tiefenbacher A, Strauss P, Brandl M and Blöschl G 2022 Pathways and composition of dissolved organic carbon in a small agricultural catchment during base flow conditions Ecohydrology & Hydrobiology 22 96–112Shang P, Lu Y, Du Y, Jaffé R, Findlay R H and Wynn A 2018 Climatic and watershed controls of dissolved organic matter variation in streams across a gradient of agricultural land use Science of The Total Environment 612 1442–53
Citation: https://doi.org/10.5194/hess-2023-292-RC2 - AC2: 'Reply on RC2', salmon jordy, 12 Apr 2024
Status: closed
-
RC1: 'Comment on hess-2023-292', Anonymous Referee #1, 19 Feb 2024
Review on Improving the internal hydrological consistency of a process-based solute-transport model by simultaneous calibration of streamflow and stream concentrations
In their manuscript, Salmon-Monviola et al. explore the utilization of dissolved organic carbon (DOC) and nitrate concentrations as constraints to refine streamflow predictions and enhance the internal consistency of a conceptual hydrological model. While their investigation revealed that DOC and nitrate concentrations did not enhance streamflow predictions per se, they did, however, reduce the uncertainty in model parameters and the representation of internal hydrological states and flow. The manuscript is notably well-written and clearly makes a case for improving the internal consistency of the conceptual hydrological models by adding additional constraints, such as solute concentrations. It convincingly illustrates how the inclusion of DOC and nitrate concentrations considerably affects the representation of underlying hydrological states and flow. Nonetheless, I have some doubts about whether the constrained version reflects greater realism, as it improved the representation of groundwater storage but showed equal to even worse results for soil moisture (more details in the general comments below). With these concerns addressed, the manuscript holds promise as a significant contribution to the readership of HESS.
General comments:
I find it convincing that the simulations, constrained by nitrate concentrations (S3 and S4), have improved the representation of groundwater levels. However, I cannot entirely follow the interpretation of an improved representation of upslope soil moisture. The authors reference Figure 11 to support this claim, yet upon examination, I observe only marginal disparities between the non-constraint simulation (S1) and the one constrained by DOC (S2), while S3 and S4 exhibit notably lower performance in terms of NSE, KGE, PBIAS, and RSME. Consequently, from my perspective, none of the simulations incorporating DOC and/or nitrate concentrations (S2 – S4) consistently elevate internal model consistency in representing soil moisture and groundwater. Hence, it remains uncertain whether these simulations merely exert a general influence on the representation of hydrological states and flows or, indeed, foster an overarching enhancement in the models' internal consistency. It might be that the representation of DOC and nitrate processing and transport are too simple, or soil moisture measurements are not presentative for the entire catchment, as nicely discussed in chapters 4.2 and 4.3. Nevertheless, I find this point insufficiently addressed.
I missed a discussion on the applicability across different catchments, especially in the view that dominant sources and pathways of DOC and nitrate concentrations can strongly vary in other settings. Based on my interpretation of Figures 5 and 6, it seems that DOC concentrations exhibit an enrichment pattern (i.e., increasing concentrations with increasing streamflow), whereas nitrate concentrations demonstrate dilution patterns (i.e., decreasing concentration with increasing discharge). However, these patterns may differ significantly in other catchments with distinct sources and pathways (e.g., Winter et al., 2021; Knapp et al., 2020). How might these differences affect your model setup? From my perspective, it is crucial to address the implications of the specific catchment characteristics and the transferability of your findings to other catchments.
Specific comments:
Line 95 - 96: What exactly do you mean by substantially differing? Is there a directional relationship between concentrations and discharge, or are dynamics completely independent?
Line 98: I recommend toning down a little to “can” be closely related.
Line 98 – 104: See here my second general comment. There are various patterns of DOC nitrate export dynamics depending on the storage and flow paths within the catchment. This is a little too simplified for my taste.
Line 105 – 108: Nice and clear!
Line: 116: Could you write the full name of AgrHyS once, please?
Line 135: Why only in riparian-zone soils and not in all soils? What dynamic do you infer if speaking of unlimited supply? Chemostasis? Enrichment? Information about DOC sources and the relationship between Q and DOC, and Q and nitrate (for example in the SI) could help to back up this argument.
Line 379: Those deep infiltration losses appear a little bit like a “mathematical marionette” to me, and they appear to be a highly sensitive parameter. Can you please elaborate on those a little more?
Line 454 – 457: This sentence is very hard to read. Can you rewrite or slit it, please?
Line 527 - 529: See my first major comment. The improvement in groundwater storage is convincing, but I do not see a significant improvement in upslope soil moisture. Here, S1 performs similarly well to S2 and clearly better than S3 and S4. I find it critical that S1 seems to perform better for soil moisture, while S3 and S4 perform better for groundwater. Thus, neither DOC nor NO3 seems to consistently improve the internal representation of water fluxes.
Line 541 – 543: Consequently, this appears overstated to me. I only see that it improves the groundwater storage representation, while two out of three scenarios show a lower performance for soil moisture.
Line 556 – 575: I agree with all points discussed here, but I am not entirely convinced that adding DOC and NO3 made the model produce the ‘right answers for the right reason’ for the reasons mentioned above.
602 – 604: Good point. I also enjoyed reading chapter 4.2 and 4.3.
Line 667 - 672: This appears contradictory to me. Did it or did it not improve the models' ability to reproduce streamflow? The first sentence says no, the second yes.
Figures
Figure 1: Why show the entire Narzin catchment if your results focus on the Kervidy-Narzin catchment only? In the caption, you do not mention the Narzin catchment either – It seems more straightforward to me to show the Kervidy-Narzin catchment only.
Figures 4, 5, and 6: Adding the scenario names (S1 – S4) to the figures would improve clarity compared to only mentioning them in the description below. Moreover, I find it difficult to see the differences between the model runs. You might consider adding an observed vs. simulated plot with a 1:1 line to the SI. But this is just a suggestion.
Figure 9: consider using a white font in front of the dark blue box for S1. A black font is hard to read with dark background (this also applies to the other figures).
Figures 9, 10, and 11 could be combined and reduced in height to save space.
Overall, I do think 12 Figures, each of them taking up around half a page, is too much. This might be a matter of taste, but I would recommend combining figures and lowering their height and/or shifting some of them to the SI.
References
Knapp, J. L., Freyberg, J. von, Studer, B., Kiewiet, L., and Kirchner, J. W.: Concentration-discharge relationships vary among hydrological events, reflecting differences in event characteristics, Hydrol. Earth Syst. Sci. Discuss., 1–27, https://doi.org/10.5194/hess-24-2561-2020, 2020.
Winter, C., Lutz, S. R., Musolff, A., Kumar, R., Weber, M., and Fleckenstein, J. H.: Disentangling the impact of catchment heterogeneity on nitrate export dynamics from event to long‐term time scales, Water Resour. Res., 57, e2020WR027992, https://doi.org/10.1029/2020WR027992, 2021.
Citation: https://doi.org/10.5194/hess-2023-292-RC1 - AC1: 'Reply on RC1', salmon jordy, 12 Apr 2024
-
RC2: 'Comment on hess-2023-292', Anonymous Referee #2, 06 Mar 2024
The study by Salmon-Monviola et al. investigates the utility of using biogeochemistry parameters to improve a conceptual hydrological model's accuracy and internal consistency. The authors modeled hydro-biogeochemical processes in the Kervidy-Naizin catchment in NW France to test this idea. They used a hydrological model based on previously published models, e.g., Hrachowitz et al. (2015). Adding DOC and NO3- processes into the model did not improve streamflow prediction. However, adding solute processes to streamflow improved the model's internal consistency, as demonstrated by the ability to model groundwater level and upslope soil moisture and reduce parameter uncertainty. The paper was generally well organized and well written, and the aims of the paper are within the scope of HESS. The paper requires some revisions, but if done satisfactorily, this paper would be a solid addition to the literature.
General comments:
I agree with Reviewer 1's ‘General Comments’ section. Figure 11 does not provide strong enough evidence to suggest that adding DOC to the model (S2) improves model (S1) accuracy. I would also like to echo the need for discussion on the applicability of these findings across regions.I also found it difficult to follow how results were produced and interpreted. This may be because I don’t have experience with this type of modeling. However, HESS has a broad hydrological readership, so the following comments are aimed to help communicate to a broader audience. For example, it wasn't easy to follow how the DOC and NO3 parameters were incorporated into the conceptual model. The hydrology was well outlined in Figure 2, and it would help to have a similar figure (perhaps Fig. 2b) for the biogeochemistry. Secondly, many model performance metrics (e.g., PBIAS, RMSE, eCDF, etc.) were not adequately explained or introduced, making it challenging to follow model interpretations. Additionally, the results alluded to some statistics (e.g., line 515), but no statistical methods were described. Another inconsistency was in the figures. Figures 4-6 had distinct calibration and evaluation periods for streamflow, DOC, and NO3 concentrations. Yet for figures 9-11 showing groundwater level and soil moisture, the whole period appeared to be for evaluation, but the date ranges on the x-axis varied. It wasn’t clear how these modeling approaches differed and why the date ranges changed.
Finally, as a stylistic point, I found all the model abbreviations and acronyms in the text highly distracting. I suggest that to improve readability, the authors should only use common abbreviations (e.g., DOC) in the text.
Specific comments:
Lines 87-88: vague sentence. The example needs to be more concrete.
Lines 95-97: What is “This potential” referring to? The spatial distribution of solutes where? In groundwater?
Line 130: are ‘livestock units’ different from the number of animals?
Line 157: define TDR
Lines 219-223: need more clarification in this section. Are you suggesting more N is removed in winter via denitrification, and in summer by biological uptake?
Lines 232-233: The two Birkel et al. papers cited here are from catchments with wetlands supplying the bulk of the DOC. So, it isn’t surprising that the stream DOC reactivity is negligible. However, DOC is typically more reactive in agricultural catchments (see (Shang et al 2018, Eder et al 2022)). However, it doesn’t seem like stream water is a reservoir with an associated transit time in the conceptual model, so how would stream water DOC (and NO3) reactivity be included? It seems like your model is well-positioned to speak to solute groundwater transport, but less so for surface water.
Lines 235-236: I think this is a fair assumption, but I would reword the justification to say that deeper mineral soils are DOC sinks.
Line 240: why is the ‘L’ in [M L-3] cubed? What does the L stand for? It seems to represent units of volume, but it wouldn’t make sense for that to be cubed. Please clarify and be consistent throughout the text.
Lines 281-288: two very long sentences. Consider revising it into multiple sentences.
Line 322: start new paragraph.
Line 324: please define which 6 metrics were used.
Line 345: define ‘Pareto front’ and how to interpret.
Line 393: the phrase “relatively well” is vague. Be more specific.
Lines 401-403: Consider reporting the mean and standard deviation of observed DOC and NO3 concentrations to aid RMSE interpretation. In some catchments, mean DOC and NO3 concentrations are less than or equal to your observed RMSE, so it’s important to contextualize this information.
Lines 459-461: Great sentence!
Line 478: move your key result to the beginning of this paragraph.
Line 490: move your key result to the beginning of this paragraph.
Line 491: confusing use of ‘well’—is it an adverb or noun here?
Table 2: In the DOC concentration row, Definition column, remove the word ‘rate’ in “DOC concentration rate…”
Figure 4: the way this is plotted, it’s difficult to tell when the observed and simulated data are just the same or there are gaps in the data.
Figure 6: when the model predicts highest annual concentrations there is no observational data. What happened to that data? Is it missing? This needs to be explained in the methods, and some discussion is needed to explain how this might affect model accuracy.
Figure 7: the last sentence of the caption states that “The boxplot of KGENO3 for scenarios S1 and S2 are absent because their values were negative.” What do negative values mean for model performance?
Figure 8: text is too small in the graphs. It might also be helpful to explain how to interpret the eCDF values somewhere in the manuscript.
Figures 9-11: The background color in the table insert makes the text illegible. Also, font size is too small.References
Eder A, Weigelhofer G, Pucher M, Tiefenbacher A, Strauss P, Brandl M and Blöschl G 2022 Pathways and composition of dissolved organic carbon in a small agricultural catchment during base flow conditions Ecohydrology & Hydrobiology 22 96–112Shang P, Lu Y, Du Y, Jaffé R, Findlay R H and Wynn A 2018 Climatic and watershed controls of dissolved organic matter variation in streams across a gradient of agricultural land use Science of The Total Environment 612 1442–53
Citation: https://doi.org/10.5194/hess-2023-292-RC2 - AC2: 'Reply on RC2', salmon jordy, 12 Apr 2024
Model code and software
Modeling water, and nitrate and dissolved organic carbon concentrations dynamics in an agricultural headwater catchment Jordy Salmon-Monviola, Ophélie Fovet, and Markus Hrachowitz https://doi.org/10.5281/zenodo.10161243
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 442 | 172 | 38 | 652 | 27 | 36 |
- HTML: 442
- PDF: 172
- XML: 38
- Total: 652
- BibTeX: 27
- EndNote: 36
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1