Wastewater matters: Incorporating wastewater reclamation into a process-based hydrological model (CWatM v1.08)

Fridman, Dor; Smilovic, Mikhail; Burek, Peter; Tramberend, Sylvia; Kahil, Taher

doi:https://doi.org/10.5194/gmd-2024-143

Preprints

https://doi.org/10.5194/gmd-2024-143

Preprints

Submitted as: development and technical paper

02 Aug 2024

Submitted as: development and technical paper |

| 02 Aug 2024

Status: this preprint is currently under review for the journal GMD.

Wastewater matters: Incorporating wastewater reclamation into a process-based hydrological model (CWatM v1.08)

Dor Fridman, Mikhail Smilovic, Peter Burek, Sylvia Tramberend, and Taher Kahil

Abstract. Wastewater treatment and reuse are increasingly perceived as essential to improve water use efficiency and increase water availability and reliability. Furthermore, wastewater has a significant impact on hydrological signals in urban watersheds. Hydrological modeling has developed over the last few decades to account for the human-water interface. Yet, despite the importance of wastewater treatment and reclamation, it is not yet comprehensively included in large-scale and multi-resolution hydrological models. This paper presents the newly developed wastewater treatment and reclamation module as part of the hydrological Community Water Model (CWatM) and demonstrates its capabilities and advantages in an urban and watershed with intermittent flows. Incorporating wastewater into the model increases model performance by better-representing discharge during the dry period. It allows for representing wastewater reuse in different sectors and takes on a modular approach, allowing for higher control over the wastewater treatment and reclamation process when spatial resolution and data availability allow it. As the current development focuses on water quantity, the water quality dimension of wastewater treatment remains a limitation, which sets the plans of incorporating water quality into the model and developing global input data for wastewater treatment and reclamation.

Received: 23 Jul 2024 – Discussion started: 02 Aug 2024

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 preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Dor Fridman, Mikhail Smilovic, Peter Burek, Sylvia Tramberend, and Taher Kahil

Status: open (until 27 Sep 2024)

Post a comment Subscribe to comment alert

CEC1:
'Comment on gmd-2024-143', Juan Antonio Añel, 14 Aug 2024 reply

Dear authors,

Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".

https://www.geoscientific-model-development.net/policies/code_and_data_policy.html

First, you have archived your code on GitHub. However, GitHub is not a suitable repository for scientific publication. GitHub itself instructs authors to use other alternatives for long-term archival and publishing. Therefore, please, publish your code in one of the appropriate repositories, and reply to this comment with the relevant information (link and and permanent identifier for it (e.g. DOI)) as soon as possible, as we can not accept manuscripts in Discussions that do not comply with our policy. Therefore, the current situation with your manuscript is irregular.
Secondly, the address "https://github.com/dof1985/CWatM-Israel" does not even work.
In this way, if you do not fix these problems, we will have to reject your manuscript for publication in our journal.
Also, you must include in a potentially reviewed manuscript the modified 'Code and Data Availability' section, the DOI of the code (and another DOI for the dataset if necessary).

Juan A. Añel

Geosci. Model Dev. Executive Editor

Reply

Citation: https://doi.org/10.5194/gmd-2024-143-CEC1
- AC1:
  'Reply on CEC1', Dor Fridman, 20 Aug 2024 reply
  Dear Juan A. Añel,
  
  We thank you for your comment and would like to apologize for this error. We have now fixed these shortcomings and aligned the model and data availability section with the journal policy. To address your specific comments we have taken the following steps:
  
  The CWatM-Israel model was made public (instead of private) and a 'frozen' copy of the version used for this manuscript is stored in Zenodo under: https://doi.org/10.5281/zenodo.13324062
  
  The Data is available under https://doi.org/10.5281/zenodo.12752967
  
  The module is integrated in the development branch of the model and would be available as a model release when/if the manuscript is accepted (https://github.com/iiasa/CWatM/tree/develop).
  
  Based on communications with the editorial support, we are not able to update the preprint file (Please refer to the response below). Therefore we have provide below the revised text of the code and data availability section. We are further able to share a revised version of the manuscript upon request.
  
  We hope these modifications meet the guidelines and policy of the journal.
  
  On behalf of the authors,
  
  Dor Fridman
  
  **** Modified text:
  
  The results presented in this manuscript used the CWatM-Israel v1.06 (last accessed: 14/08/2024; https://doi.org/10.5281/zenodo.13324061), and the Israel 1km input data for CWatM (last accessed: 17/08/2024; https://zenodo.org/doi/10.5281/zenodo.12752966). The inclusion of the wastewater treatment and reclamation into CWatM is available under the development branch (last accessed: 17/08/2024; https://github.com/iiasa/CWatM/tree/develop).
  
  **** Editorial support comment:
  
  Dear Dor Fridman,
  
  Many thanks for your email.
  
  Please need to change/add this to your revised version, as we cannot replace the preprint. You can post a reply to inform the chief editor about this.
  
  Kind regards,
  
  Anna
  
  Reply
  
  Citation: https://doi.org/10.5194/gmd-2024-143-AC1
  - CEC2: 'Reply on AC1', Juan Antonio Añel, 20 Aug 2024 reply
    
    Dear authors,
    
    Unfortunately, your reply does not satisfactorily address the issues that I pointed out, and under these conditions, we must reject your manuscript for publication.
    We requested that you store all the code in permanent repositories that comply with our policy. However, in your reply, you continue to fail to address this issue, failing to share the module developed and presented in your manuscript according to our policy. You state that it will be available as a "model release". We can not accept such conditions. As I mentioned in my previous comment, and it is evident in the policy (which I linked), all the code must be permanently stored before submitting the manuscript.
    Therefore, I urge you once more to promptly and correctly store all the code used in your manuscript in accordance with our policy. Failure to do so will result in the suspension of discussions on your manuscript and its subsequent rejection. It is crucial to understand that, for the sake of fairness to other authors, reviewers, editors, and the scientific publication process, we can not afford to invest time and resources in a manuscript that does not adhere to the rules for scientific publication.
    Juan A. Añel
    Geosci. Model Dev. Executive Editor
    
    Reply
    
    Citation: https://doi.org/10.5194/gmd-2024-143-CEC2
    
    AC2: 'Reply on CEC2', Dor Fridman, 20 Aug 2024 reply
    
    Dear Juan A. Añel,
    We fully acknowledge the importance of adhering to the journal’s rules and policies, including the provision of data and code to ensure reproducibility. Our commitment to this standard is evident in the series of CWatM papers already published in GMD (e.g., https://doi.org/10.5194/gmd-17-5123-2024 and https://doi.org/10.5194/gmd-16-2437-2023).
    With respect, we differ in the observation that our submission does not comply with the 'Code and Data Policy.' We have ensured that both the code and dataset are accessible on Zenodo, utilizing the Zenodo GitHub integration as recommended by the journal. Additionally, we have provided DOIs for both the code and dataset.
    The reference to the model on GitHub was included solely to indicate that the wastewater treatment and reclamation module has been integrated into the main version of the Community Water Model, which is widely utilized by the community. However, this version of the model was not employed to generate the results presented in this manuscript. For the sake of clarity, we are amenable to removing this additional GitHub reference, as it is not necessary for reproducing the results. Therefore, we respectfully propose the following revised paragraph:
    The complete model (CWatM-Israel v1.06) used to conduct the simulations presented in this manuscript is available from https://doi.org/10.5281/zenodo.13324061 (last accessed: 14/08/2024). The input data used for this publication can be downloaded from https://zenodo.org/doi/10.5281/zenodo.12752966 (last accessed: 17/08/2024). The Community Water Model (CWatM) manual can be accessed via https://github.com/iiasa/CWatM.
    
    Reply
    
    Citation: https://doi.org/10.5194/gmd-2024-143-AC2
    
    CEC3: 'Reply on AC2', Juan Antonio Añel, 20 Aug 2024 reply
    
    Dear authors,
    I am sorry to be so direct. I do not understand if you are unclear on the purpose or if I explain it so badly. A question remains: Is the module that you present in your manuscript included in the Zenodo repository that you now provide? If yes, where exactly? That's the only outstanding issue. I want to understand that it is included, but from your first reply to my comment and the last one, it remains unclear.
    
    Also, could you remove all the GitHub references that create a mess when trying to assess the compliance of your manuscript with our policy? You can include the manual in the Zenodo repository, too, to provide readers with it. It is important that you understand that the Code and Data Availability section exists to assure the replicability of the manuscript and compliance with the scientific method of the research presented, not for promotion. It is of GitHub, a project webpage, or future versions of a model.
    We would thank a clear reply with only the permanent repositories (DOI and link) for the full model version used in your manuscript, the new module presented (or identification of where it is located in the tree of directories of the model previously mentioned), and the data.
    Juan A. Añel
    Geosci. Model Dev. Executive Editor
    
    Reply
    
    Citation: https://doi.org/10.5194/gmd-2024-143-CEC3
    
    AC3: 'Reply on CEC3', Dor Fridman, 21 Aug 2024 reply
    
    Dear Juan A. Añel,
    The module presented in the paper is fully integrated into the model provided in shared links (see below). It is not included in one script, but is integrated in multiple files of CWatM. We provide the filenames + location of the main model components, but we do not think it should be in the code and data section - as one who would like to replicate this work would not have to access these files.
    The files are: 'cwatm/hydrological_modules/water_demand/wastewater.py'; 'cwatm/hydrological_modules/water_demand/water_demand.py'
    We propose this paragraph for the code and data section:
    The ‘wastewater treatment and reclamation module’ is fully-integrated into the Community Water Model (CWatM). The model version used to conduct the simulations presented in this manuscript (CWatM-Israel v1.06) is available from https://doi.org/10.5281/zenodo.13324061 (last accessed: 14/08/2024). The input data used for this publication can be downloaded from https://zenodo.org/doi/10.5281/zenodo.12752966 (last accessed: 17/08/2024).
    
    Reply
    
    Citation: https://doi.org/10.5194/gmd-2024-143-AC3
    
    CEC4: 'Reply on AC3', Juan Antonio Añel, 21 Aug 2024 reply
    
    Dear authors,
    
    Thanks for the clarification. I think that your manuscript is now in compliance with our policy. However, given that your "module" is not standalone, and, it is my understanding, not software that can be dynamically loaded but code added on top of existing scripts, it would be good to provide a better identification of the code you have implemented. The diff files of the model without your new code and with it would be the best way to identify the latest code in your submission.
    
    In the same vein, it would be better not to use the word "module" to refer to the new code in the manuscript to avoid misunderstandings.
    
    Regards,
    
    Juan A. Añel
    
    Geosci. Model Dev. Executive Editor
    
    Reply
    
    Citation: https://doi.org/10.5194/gmd-2024-143-CEC4
RC1:
'Comment on gmd-2024-143', Anonymous Referee #1, 02 Sep 2024 reply
The study introduces and describes a new “wastewater reclamation module” for the global hydrological model CWatM, evaluating the impact of the new module from a water quantity perspective by comparing observed and simulated river discharge for a case study location (Ayalon basin). This is an important area of research for (large-scale) hydrological models, with both aspects of water quantity and quality associated with wastewater have been largely overlooked and simplified in existing models. However, I find it somewhat difficult to fully follow the approach taken, and therefore I find that the manuscript would benefit greatly from a more comprehensive description of the three stages included in the workflow (i.e. pre-treatment; treatment; post-treatment), in addition to justifications for the assumptions made.
For example:
Line 115: “User-defined collection areas”. What exactly does this mean/ how is the service area of a wastewater treatment plant defined? How applicable is the methodology for applying in areas lacking detailed information on wastewater treatment plants (i.e. export share, designed HRT)? While not tackled directly in this paper, the authors state that this work “sets the plans of “developing global input data for wastewater treatment and reclamation”, but it seems to rely on plant-specific information that is not always readily available.

How is seasonality in produced wastewater (and therefore reclamation) accounted for in the model? Line 127 suggests daily influent versus daily treatment capacity is considered; but does produced wastewater from e.g. domestic and industrial water uses actually vary at that temporal resolution? Similarly, in some areas, treated wastewater reuse may only occur intermittently throughout the year. How would the model deal with this/ is there a scheme for allowing temporal variability in wastewater re-use?

Line 127: “excess wastewater is discharged into pre-specified discharge locations”. What exactly does this mean, and how are these specified? Are these specified discharge locations storage basins, or the surface water network itself (i.e. overflows)?

Line 140: How is the “surface area of treatment pools” estimated? Why is the estimated pool depth set to 6m? Then line 153 says the depth is set to 1m (does it differ for extensive and intensive systems; it is not clear)? Also, what is the rationale for assuming three treatment pools?

Line 173: “collected untreated wastewater is exported form the simulated region if the WWTP associated with the collection area does not exist”. Please elaborate, I do not understand what is meant here.

Line 178: How is it determined if a reservoir is accepting treated wastewater? What is the data source for reservoirs and their “associations” with wastewater treatment plants?

The case study and scenarios are interesting. However, the model is applied and validated against observed discharge in only a single basin; which is both water scarce and is already large reliant on treated wastewater re-use. I find it therefore difficult to assess the applicability of the proposed approach in other hydrological conditions and with different levels of data availability. Being a module of a global hydrological model, I can imagine the eventual intention is to scale up this to be applicable for modelling the globe. I understand the push towards a multi-resolution modelling framework, but I currently struggle to see how/if this wastewater module would be implemented at more coarse spatial resolutions and in more data poor regions.

Reply
Citation: https://doi.org/10.5194/gmd-2024-143-RC1
RC2:
'Comment on gmd-2024-143', Anonymous Referee #2, 06 Sep 2024 reply
In this manuscript, the authors introduce a new module that considers wastewater treatment and reclamation (WRTM) to the Community Water Model (CWatM). Additionally, the manuscript also provides model performance analysis of multiple scenarios with and without the new module as well as some additional wastewater balance analysis for a case study location (Ayalon basin). In general, the manuscript targets wastewater processing, which is an important aspect of research in large-scale hydrological modeling that is often overlooked. However, similar to RC1, I believe the manuscript still needs major revisions before it can be considered for publication.

Major comments:
Module variables: while there are brief explanations of each variable following equations, I find some of those explanations confusing. For example: line 109, “…a logical variable (e.g., Ddom).”, what do you mean by logical? Or line 110, “…share coefficient (Cs) representing sewer connection rates and leakages”, so does that mean if Cs is set at 100, the sewer system will have 100% connection with no leakage? Considering that this module is intended to be publicly used, it would be beneficial to both the manuscript and readers if the authors would provide a much more detailed description of the variables in Equ.1 on: (1) How can they be identified? (2) What are the ranges and units (if any) of each variable? And (3) What do specific values of each variable mean? Please consider providing additional sensitivity analysis on these parameters as well.

Module inputs: Table 1 has provided a summary of the model’s input; however, it is unclear to the reader what the specific temporal (hourly, daily, or monthly? Timeseries or fixed value?) and spatial (gridded or vector-based?) requirements of the local datasets which hindered the possibilities of replicating this manuscript results or future applications of new model users. Additionally, it is unclear which input is critical and which is optional making it challenging to consider applying the module in other data-limited regions.

Model calibration: results from Table 3 suggest that while there are significant improvements when the module is applied, the simulated dry season mean flow is still substantially higher than observed data. Additionally, results in Fig B1 suggest that simulated results overestimate ET substantially across the entire calibration period (line chart) while normalized differences are also high, where the majority of the basin in spring and many locations in summer have ~70 to >100%. Thus, please consider further improving the model performance and provide additional comparisons for the validation period as well.

Model validation: considering that this paper focuses on introducing a new module, I’m struggling to see a clear comparison of model performance before and after the module is applied. It is rather difficult to distinguish the difference between simulated and observed results in Figure 3, thus, please consider providing an additional figure showing the hydrographs in a shorter period (i.e., 1 year) and displaying results from all scenarios so that the comparison is more visually clear how the new module can improve model performance. Additionally, since wastewater inflow is also a factor in evaluating model performance, please consider adding additional statistical values in Figure 5 (i.e., NSE, KGE).

Study area: similar to RC1, I find it difficult to assess the applicability of the module with just validation of one basin/one gauging station. I’d suggest the authors consider adding additional analysis on other basins where there is either less or more data available.

Minor comments
Title, I believe the new module not only considers wastewater reclamation but also additional treatment time right, perhaps a different title that contains WRTM?

Line 91, “hyper-spatial resolution”, while 1km grid is relatively high-resolution compared to most global models, please consider editing this term across the manuscript since there are other small-scale models that can simulate wastewater process at much higher resolution.

Line 87, “MODFLOW6” and line 226, “Modflow”, are these two the same? If so, please be consistent across the manuscript.

Figure 2, currently, the legend is rather difficult to read with all the visualization mixed in. Please consider turning all areas outside of the study area to white color in the main map to make the figure more visible.

Figure 4, in-figure text size is too small, please consider increasing it.

Line 281-284, “… simulated evapotranspiration with a satellite derived product… the simulated monthly influent flows into the Ayalon WWTP with observed data…”, please reference these figures.

The “detrended values” shown in Figure 5 are not mentioned specifically anywhere in the main text, thus consider changing the text to have more linkage while adding more detailed information on the detrend technique. Additionally, there seems to be a clear increasing trend in the top panel of Figure 5, which might be eliminated if the authors consider running an additional spin-up period to stabilize the model.

All tables, table sizing, and alignments are inconsistent, please revise.

Reply
Citation: https://doi.org/10.5194/gmd-2024-143-RC2
RC3: 'Comment on gmd-2024-143', Anonymous Referee #3, 07 Sep 2024 reply

The manuscript titled “Wastewater matters: Incorporating wastewater reclamation into a process-based hydrological model (CWatM v1.08)” introduces a novel development by integrating a wastewater treatment and reclamation module into the CWatM model. This advancement is important as it incorporates human-related activities into large-scale hydrological modeling. The overall structure of the paper is sound and logically organized. However, several aspects of the methodology require further detail and clarification. Additionally, there are concerns regarding the model calibration and validation design, which could influence the results and their discussion. I recommend major revisions to address the methodology section and enhance the robustness of the model evaluation.

Detailed comments
Line 89: Please verify whether the acronym "WRTM" is correct or if it is a typographical error. The acronym for "Wastewater Treatment and Reclamation Module" should be "WTRM" rather than "WRTM". If it is a typo, please correct it throughout the manuscript.
Lines 122-124: Please clarify the criteria for defining a wastewater treatment plant (WWTP) with a hydraulic retention time (HRT) ranging from 24 hours to 2 days. Would it be intensive or extensive?
Lines 127-130: The description here is unclear. Line 127 states that if the influent exceeds the designed capacity, the excess wastewater is discharged to a pre-specified location. This implies that the inflow is managed below designed capacity by discharging excess wastewater. However, Line 128 mentions that treatment plants allow inflows to exceed the designed capacity, which appears contradictory. Please clarify whether "daily treatment capacity" and "designed capacity" refer to the same or different metrics.
Line 136-137: Please clarify what “an increase of 25% in the operational daily capacity” is compared to? E.g., compared to HRT = 1 day?
Lines 154-155: The term "total treatment time" is ambiguous. Is this time fixed for treating fully filled treatment pools, or does it vary based on the storage conditions? Please provide a more detailed explanation.
Line 180: How are WWTPs assigned to "associated reservoirs"? For instance, must the reservoir be downstream of the WWTP, or can users assign reservoirs to WWTPs freely? Please clarify.
Lines 266-267: Please clarify whether Scenario S2 was used for model calibration and if the calibrated parameters were applied across all scenarios. If so, this approach might not be appropriate, particularly in the context of the model performance evaluation discussed in Section 4.1 (e.g., Table 3). Scenarios S0 and S1 may underperform because they were not calibrated for these specific scenarios. Please explain why separate calibration for each scenario was not performed.

Figures
Figure 2: The reservoirs, canals, and other water features are difficult to distinguish due to the busy map. Consider increasing the transparency of the background map, removing it altogether, or using thicker lines to better highlight the water features.
Figure 4: The figure is confusing, as it is unclear which text corresponds to which circle. It would be helpful to use different colors for the text associated with the circles and the text indicating different flows to improve clarity. In addition, the texts in the circles are too small to see.

Reply

Citation: https://doi.org/10.5194/gmd-2024-143-RC3

Dor Fridman, Mikhail Smilovic, Peter Burek, Sylvia Tramberend, and Taher Kahil

Viewed

Total article views: 498 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
326	99	73	498	5	6

HTML: 326
PDF: 99
XML: 73
Total: 498
BibTeX: 5
EndNote: 6

Views and downloads (calculated since 02 Aug 2024)

Month	HTML	PDF	XML	Total
Aug 2024	226	78	26	330
Sep 2024	100	21	47	168

Cumulative views and downloads (calculated since 02 Aug 2024)

Month	HTML	PDF	XML	Total
Aug 2024	226	78	26	330
Sep 2024	100	21	47	168

Viewed (geographical distribution)

Total article views: 535 (including HTML, PDF, and XML) Thereof 535 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Sep 2024

Short summary

Global hydrological models are applied at high spatial resolutions to quantify water availability and evaluate water scarcity mitigation options. Yet they mostly oversee important local processes. This paper presents and demonstrates the inclusion of wastewater treatment and reclamation into a global hydrological model. As a result model performance is improved, and models are capable to utilize treated wastewater as an alternative water source.


Total:	0
HTML:	0
PDF:	0
XML:	0