Monitoring Groundwater Use as a Domestic Water Source by Urban Households: Analysis of Data from Lagos State, Nigeria and Sub-Saharan Africa with Implications for Policy and Practice

Round 1

Reviewer 1 Report

Comments on:

“Monitoring groundwater use as a domestic water source by urban households: Analysis of data from Lagos, Nigeria and sub-Saharan Africa with implications for policy and practice” (water-1080487)

The manuscript entitled “Monitoring groundwater use as a domestic water source by urban households: Analysis of data from Lagos, Nigeria and sub-Saharan Africa with implications for policy and practice” examines existing databases informing on domestic groundwater use in sub-Saharan Africa.

I think that there is an important effort of data recompilation and analysis, and that the analysis done has the potential to help in water policy design. However, I think that this analysis needs to be strengthened (now, it is a simple description of databases), and that there are other major issues that should be solved before publishing this paper.

Major comments:

1. Authors need to be more concise. The paper is too long and, in some cases, excessive explanations or details are provided. Some examples are Box 1 (that can be eliminated, and substituted only by the UN reference), or the explanation of lines 128-138 (an argument that is well known by the Water Journal readers). These are only two examples, but a synthesis needs to be done in all the manuscript.
2. In Sections 2, 3.2., and 5.2. I would recommend eliminating subsections because they hamper the readability of the text. Arguments needs to be more linked between them.
3. In Section 2, a more detailed review about groundwater use drivers needs to be done. There is an unneglectable literature on this topic than needs to be reviewed and discussed in this section. A good review of this point could be one of the strongest points of the paper and could help the authors to stablish how the existing databases can provide information (or not) to address this sort of analysis. Some of the papers I would recommend to review are:

• Cheesman, J.; Bennett, J.; Son, T.V.H. Estimating household water demand using revealed and contingent behaviors: Evidence from Vietnam. Water Resour. 2008, 44, W11428.
• Coulibaly, L.; Jakus, P.M.; Keith, J.E. Modeling water demand when households have multiple sources of water. Water Resour. Res. 2014, 50, 6002–6014.
• Nauges, C.; Strand, J. Estimation of non-tap water demand in Central American cities. Resour. Energy Econ. 2007, 29, 165–182.
• Nauges, C.; van der Berg, C. Demand for piped and non-piped water supply services: Evidence from Southwest Sri Lanka. Environ. Resour. 2009, 42, 535–549.

Moreover, some reference needs to be done in this section to the existing literature on industrial groundwater use. Although this literature is more focused on developed countries, competition for groundwater between household and industrial uses could be a source of aquifer oversploitation. This issue needs to be addresses to contextualise the importance of policies for groundwater management. Some relevant papers are:

• Reynaud, A. An econometric estimation of industrial water demand in France. Resour. Econ. 2003, 25, 213–232.
• Gracia-de-Rentería, P.; Barberán, R.; Mur, J. The groundwater demand for industrial uses in areas with Access to drinking publicy-supplied water: a microdata analysis. Water. 2020, 12, 198.
• Arbués, F.; García-Valiñas, M.A.; Villanúa, I. Making decisions on industrial water sources: the case of Zaragoza, Spain. Urban Water Journal. 2020, 17(2), 122-135.

4. The analysis of the databases in Sections 3.2., 4.2., and 5.2. need to be improved. This is my main concern about the paper, since now it is a simple description of the databases results. Moreover, I have detected some possible data errors in the presentation of the results (for example, MICS data from Table 2 do not coincide with data from Table WS1 of MICS report). MICS data also provide information by education level, so it should be also analyse as a possible driver of groundwater use.
5. I think that in the discussion section, the potential of the existing data should be discussed. Do these data allow to carry out empirical analysis such as the developed by the existing literature for other countries? Could be the groundwater use drivers been estimated for the case study?
6. Authors should be more cautious with some of the claims they make in the discussion section. For example, in lines 965-967, the authors said “our analysis indicates the role that wealth and income can play in mediating access to different sources of domestic water supplies, but, at least in the case of Lagos State, this may be less significant than is often stated”. And, in lines 758-760 “Interpretation of the data suggests that income is not a predictor of self-supply, but the importance of property ownership indicates that the urban poor are less likely to be able to make use of their own self-supply options”.  These statements cannot be affirmed based on the simple description of a survey. Instead, a statistical test is needed to confirm if there are statistically significant differences by income group or by rural/urban population. I would invite the authors to revise the literature to use the most suitable statistical techniques to carry out the most adequate analysis of data.

Minor issues:

1. The style of the references needs to be revised in some cases. Examples in lines 110, 119, 152, 167, 273
2. The quality of the figures needs to be improved. Moreover, the legend of figures 5 and 6 should be modified; since it could be misunderstood (it could be interpreted as the % of urban population with respect of total population).

Author Response

We wish to thank the three Reviewers for their constructive, detailed and considered comments. We are pleased that all three reviewers feel that the paper makes an important contribution to this relevant topic. We have taken the various suggestions on board, particularly the need to shorten the paper, and feel that doing so has significantly strengthened the paper. We are grateful for the time and attention each Reviewer has given to our paper and hope that we have been able to do justice to their advice. Our detailed response to individual comments follows.

Major comments:

1. Authors need to be more concise. The paper is too long and, in some cases, excessive explanations or details are provided. Some examples are Box 1 (that can be eliminated, and substituted only by the UN reference), or the explanation of lines 128-138 (an argument that is well known by the Water Journal readers). These are only two examples, but a synthesis needs to be done in all the manuscript.

We have significantly reduced the length of the manuscript. We have synthesised material in Sections 1 and 2 and removed Box 1 as suggested. We believe that the paper is now more concise and appreciate the steer towards achieving this.  We have also made significant amendments to later sections, as detailed below, which contribute to shortening the paper overall.

1. In Sections 2, 3.2., and 5.2. I would recommend eliminating subsections because they hamper the readability of the text. Arguments needs to be more linked between them.

We have eliminated sub-sections across the manuscript to make sections flow more coherently. In Section 2, we have retained three sub-sections as we felt that these helped shape the key components of the review.

3. In Section 2, a more detailed review about groundwater use drivers needs to be done. There is an unneglectable literature on this topic than needs to be reviewed and discussed in this section. A good review of this point could be one of the strongest points of the paper and could help the authors to stablish how the existing databases can provide information (or not) to address this sort of analysis. Moreover, some reference needs to be done in this section to the existing literature on industrial groundwater use. Although this literature is more focused on developed countries, competition for groundwater between household and industrial uses could be a source of aquifer oversploitation. This issue needs to be addresses to contextualise the importance of policies for groundwater management.

We welcome the suggestion to include a more detailed review about groundwater drivers and to include material on industrial groundwater use. We are grateful for the pointers to suggested sources and have drawn on this to highlight existing work on drivers of urban water demand, the nature of piped and groundwater sources as potential substitutes and the significance of industrial users as potential competitors for groundwater demand.

4. The analysis of the databases in Sections 3.2., 4.2., and 5.2. need to be improved. This is my main concern about the paper, since now it is a simple description of the databases results. MICS data also provide information by education level, so it should be also analyse as a possible driver of groundwater use.Moreover, I have detected some possible data errors in the presentation of the results (for example, MICS data from Table 2 do not coincide with data from Table WS1 of MICS report).

We have strengthened the analysis of the databases including consideration of additional drivers of demand for the development of self-supplied groundwater sources (including education, tenure and number of rooms used for sleeping). The suggestion to undertake more detailed statistical analysis was explored but we feel that the data available for urban self-supply and groundwater dependency makes this problematic in the context of the current paper.

We are grateful for the suggestion to review the data in Table 2. This Table refers to Table WQ3 (safely Managed Water Supplies) on p.69 of Lagos Report from Nigeria MICS 2016-17). The data presented is correct, but the citation suggested by the authors of this report is misleading as it refers to the National MICS Report for Nigeria. We have amended the suggested citation to explicitly reference the study for Lagos State.

5. I think that in the discussion section, the potential of the existing data should be discussed. Do these data allow to carry out empirical analysis such as the developed by the existing literature for other countries? Could be the groundwater use drivers been estimated for the case study?

We have substantially amended the discussion section. We highlight some of the weaknesses of the existing data available from the surveys analysed and reflect on how future surveys might address the gaps identified.

6. Authors should be more cautious with some of the claims they make in the discussion section. For example, in lines 965-967, the authors said “our analysis indicates the role that wealth and income can play in mediating access to different sources of domestic water supplies, but, at least in the case of Lagos State, this may be less significant than is often stated”. And, in lines 758-760 “Interpretation of the data suggests that income is not a predictor of self-supply, but the importance of property ownership indicates that the urban poor are less likely to be able to make use of their own self-supply options”.  These statements cannot be affirmed based on the simple description of a survey. Instead, a statistical test is needed to confirm if there are statistically significant differences by income group or by rural/urban population. I would invite the authors to revise the literature to use the most suitable statistical techniques to carry out the most adequate analysis of data.

Thank you for your advice here. We have taken this on board and revised our phrasing to ensure that our statements do not suggest findings that we are not able to substantiate through statistical analysis. We welcome the suggestion to undertake a more detailed statistical analysis of the available data. This has not been feasible for the present paper but is an important avenue to consider for future work and we are grateful that this has been suggested.

Minor issues:

1. The style of the references needs to be revised in some cases. Examples in lines 110, 119, 152, 167, 273
2. The quality of the figures needs to be improved. Moreover, the legend of figures 5 and 6 should be modified; since it could be misunderstood (it could be interpreted as the % of urban population with respect of total population).

We have amended the style of the references and revised the presentation of our Figures.

Reviewer 2 Report

The paper reports on a meta-analysis of different datasets that include data on groundwater uses for domestic purposes. Conclusions primarily relate to the significance of groundwater use in several regions in sub-Saharan Africa, to shortcomings in available data sets in particular, and recommendations for further research, in particular data collection.

The paper addresses a relevant topic, and is generally well-written and clear in its objectives.  

I have two major comments/suggestions.

First, the paper is lengthy, and the central part (pages 7 – 26, lines 230-640) reads more as a (sometimes detailed) report than as a paper. As a reader, I felt overwhelmed with detail and analysis data, and wondered whether I needed to go through all these details and full tables to understand the main messages of the paper. My conclusion is that the paper will become more accessible and attractive to a wider audience by significantly cutting down on the length and detail of this middle part, to the extent of reducing its length to less than 50% of the current almost 20 pages. My recommendation is therefore to be much more sparse in this middle part, focus on key observations and conclusions from the analysis, provide details/full tables only when these are necessary to understand key conclusions, and shift many tables to either appendices/supplementary materials, or refer to publicly accessible analysis reports where the material may be found.

Second, in the discussion and conclusions, the authors rightfully focus on the variety of shortcomings in the available data. However, contrary to what the last part of the title suggests, implications for policy ad practice are mentioned in very general terms only, if addressed at all.

As a result, I often felt like asking ‘so what?’ in several places where shortcomings in data are pointed out, and recommendations for more analysis/research given. What would (need to) change in policies at local, regional, national, international level if more accurate percentages were known on the extent of groundwater use, use of (private, collective) boreholes, etc. etc. ?

Providing more specific consequences (that may be different per level) of knowing better and at more detail would definitely provide stronger arguments for the recommendations given.

Author Response

We wish to thank the three Reviewers for their constructive, detailed and considered comments. We are pleased that all three reviewers feel that the paper makes an important contribution to this relevant topic. We have taken the various suggestions on board, particularly the need to shorten the paper, and feel that doing so has significantly strengthened the paper. We are grateful for the time and attention each Reviewer has given to our paper and hope that we have been able to do justice to their advice. Our detailed response to individual comments follows.

First, the paper is lengthy, and the central part (pages 7 – 26, lines 230-640) reads more as a (sometimes detailed) report than as a paper. As a reader, I felt overwhelmed with detail and analysis data, and wondered whether I needed to go through all these details and full tables to understand the main messages of the paper. My conclusion is that the paper will become more accessible and attractive to a wider audience by significantly cutting down on the length and detail of this middle part, to the extent of reducing its length to less than 50% of the current almost 20 pages. My recommendation is therefore to be much more sparse in this middle part, focus on key observations and conclusions from the analysis, provide details/full tables only when these are necessary to understand key conclusions, and shift many tables to either appendices/supplementary materials, or refer to publicly accessible analysis reports where the material may be found.

Thank you for your suggestion to reduce the length of the central part of the paper. As suggested, we have removed a number of tables and figures from the main paper and included them in the Supplementary Material. We have also rationalised other tables to focus on the key messages. Finally, we have revised the accompanying text to further focus on the key messages. We hope that our revisions have made the main messages in the paper more accessible.

Second, in the discussion and conclusions, the authors rightfully focus on the variety of shortcomings in the available data. However, contrary to what the last part of the title suggests, implications for policy ad practice are mentioned in very general terms only, if addressed at all.

This is a very valid point. We have fundamentally revised the discussion. We now make explicit reference to policy and practice implications where relevant. We hope that we now make the link between the analytical sections and the relevance to practice and policy more explicitly.

As a result, I often felt like asking ‘so what?’ in several places where shortcomings in data are pointed out, and recommendations for more analysis/research given. What would (need to) change in policies at local, regional, national, international level if more accurate percentages were known on the extent of groundwater use, use of (private, collective) boreholes, etc. etc.? Providing more specific consequences (that may be different per level) of knowing better and at more detail would definitely provide stronger arguments for the recommendations given.

We welcome the ‘so what’ question. It is a good test of the relevance of the paper. We have sought to address this more explicitly in our revision of the discussion and the conclusions. We hope that the main messages now emerge.

Reviewer 3 Report

see attached

Comments for author File: Comments.pdf

Author Response

We wish to thank the three Reviewers for their constructive, detailed and considered comments. We are pleased that all three reviewers feel that the paper makes an important contribution to this relevant topic. We have taken the various suggestions on board, particularly the need to shorten the paper, and feel that doing so has significantly strengthened the paper. We are grateful for the time and attention each Reviewer has given to our paper and hope that we have been able to do justice to their advice. Our detailed response to individual comments follows.

The manuscript is lengthy and more like a technical report or thesis but not a research paper, it should be significantly shortened, best to keep only half of its current pages.

We welcome this suggestion and we have reduced the length of the paper substantially. We have also extensively revised the paper to focus more strongly on the research elements and reduced the aspects that read like a technical report.

It is recommended to map and visualise the groundwater resource/storage in Africa using the available data sources or through satellite based remote sensing.

We recognise the importance of highlighting the existing groundwater resource. We have included references to sources that map and visualise this resource that readers can refer to. We have not included maps in the paper itself, as our focus is more strongly targeted at identifying the extent to which urban households seek to access the groundwater resource.

How reliable are the datasets or statistics used for this study? Does water resource change significantly over years.

This is a good question. Part of the purpose of the paper is to address the reliability of the published statistics. We do this through comparing the results of different surveys that have been undertaken on the same topic in the same location at similar times. This highlights some variation in the published statistics. We have considered the question of the changeability of groundwater resources over time in our literature review. This highlights the perceived importance of groundwater as a secure means of storage, but fears for its overexploitation. Our paper has not tested these wider literatures as our focus is on the extent to which households seek to access groundwaters themselves. 

We have made the minor amendments proposed.

Round 2

Reviewer 1 Report

The paper has significantly improved. In particular, the readability of the paper was significantly improved, as well as the discussion and implication of results. I miss a statistical approach to analyse these databases, but I am aware that this could be part of a future research. Anyway, I recognize that there is a huge effort of data recompilation and analysis. And the discussion of these data sources, as wells as its potential for future research, is now more present in the paper.

I point out some typos I have detected and some minor format stuff:

205. 6, l. 205. The word “two” is repeated.
206. 7, l. 266. The reference should be [58].
207. 8, l. 278. There is a line break.
208. 12, l. 372. “MICS” instead of “MICs”.
209. 12, l. 372. Census data abbreviation used in Figure 1 (CEN) should be introduced in the text in line 372.

Figure 1. The quality of the image should be improved. In the legend, the lines can be avoided (it is enough with the coloured points to identify the data sources).

Figure 3. I would suggest to use a coloured figure in order to improve the interpretation and visualization of the figure.

Author Response

Many thanks for your helpful suggestions. We have taken all of your comments on board and have revised the text accordingly. We have also amended the Figures as suggested. We are very grateful for your positive suggestions throughout this process. Thank you.

Reviewer 3 Report

This version has been improved

Author Response

We are pleased that our revisions have sufficiently improved the manuscript. Many thanks for all your constructive advice that helped us to make these improvements.