Evaluating Groundwater Storage Change and Recharge Using GRACE Data: A Case Study of Aquifers in Niger, West Africa

Round 1

Reviewer 1 Report

Computing NASA Gravity Recovery And Climate Experiment data, the manuscript remotesensing-1631708 aims to evaluate change and recharge of groundwater storage in South West Niger. Detailed objectives are exposed at the end of the Introduction (lines 167-178). They especially deal with sustainability use of groundwater and the feature of GRACE in resource management. The manuscript needs some minor changes before publication.

For the time range analysed (2002-2021, i.e. the life span of GRACE mission) the authors found that groundwater resources are not being over-utilized and could be further developed for beneficial use. Such a result should be in agreement with some previous estimates.

Study motivation is exposed in Introduction and appears to be persuasive, while the basic features of NASA GRACE data in groundwater study are appropriately indicate throughout the subsection 3.1. The authors especially emphasized one their previous paper on the “tools designed to allow groundwater managers to easily use GRACE data to estimate groundwater storage changes over time representing aquifer storage in a region” (lines 105-107 and see also subsequent lines).

As regards the estimation of groundwater recharge, the authors emphasized the Water Table Fluctuation method almost in antagonism with the chemical techniques (lines 115-155). I suggest not to stress such an issue so much. However, finally, it is underlined that a previous application of GRACE data in groundwater change estimation (Wu et alii, 2019) “did not differ considerably with the values calculated with an environmental tracer” (lines 152-153).

About the causes of the groundwater storage changes in South West Niger, only the work of Leduc et alii (1997) is cited. Do the authors know of other hypotheses, perhaps even more recent and updated on this topic?

It would be good to enter the geographic coordinates along the frame of Figure 1. Nothing else to report regarding the Section “Study area and background”.

What uncertainties can be considered for the imputed GWSa data in Figure 7? Can the authors improve the text related to this aspect?

To estimate the Annual Recharge, the authors modify the method of Wu et alii (2019) replacing well measurements with previous computed groundwater storage anomalies (lines 388-390). This is a crucial point, replacing measured values with model-derived values. Do the authors believe that the lack of direct measures may lead to errors in the assessments that are not acceptable for management?

It would improve the readability of the manuscript to put the list of abbreviations (GWSa, WFT, CMAP, CANa, and so on) at the end (see template format, Instruction for Authors).

Author Response

Reviewer 1

Computing NASA Gravity Recovery And Climate Experiment data, the manuscript remotesensing-1631708 aims to evaluate change and recharge of groundwater storage in South West Niger. Detailed objectives are exposed at the end of the Introduction (lines 167-178). They especially deal with sustainability use of groundwater and the feature of GRACE in resource management. The manuscript needs some minor changes before publication.

For the time range analysed (2002-2021, i.e. the life span of GRACE mission) the authors found that groundwater resources are not being over-utilized and could be further developed for beneficial use. Such a result should be in agreement with some previous estimates.

Study motivation is exposed in Introduction and appears to be persuasive, while the basic features of NASA GRACE data in groundwater study are appropriately indicate throughout the subsection 3.1. The authors especially emphasized one their previous paper on the “tools designed to allow groundwater managers to easily use GRACE data to estimate groundwater storage changes over time representing aquifer storage in a region” (lines 105-107 and see also subsequent lines).

Thank you for this summary.

As regards the estimation of groundwater recharge, the authors emphasized the Water Table Fluctuation method almost in antagonism with the chemical techniques (lines 115-155). I suggest not to stress such an issue so much. However, finally, it is underlined that a previous application of GRACE data in groundwater change estimation (Wu et alii, 2019) “did not differ considerably with the values calculated with an environmental tracer” (lines 152-153).

We aren’t sure how to respond to this comment. We think the text clearly indicates that the WTF method is simply an alternative method to chemical techniques and provide examples of prior studies where both methods were used. We note that chemical methods have been used successfully in our area, but that there are potential issues, the largest of which is obtaining data.

About the causes of the groundwater storage changes in South West Niger, only the work of Leduc et alii (1997) is cited. Do the authors know of other hypotheses, perhaps even more recent and updated on this topic?

We expanded the discussion in section 1.5 to include reference to several other prior studies in the region and the hypotheses presented by these studies for increasing groundwaters storage change in the region. This includes a comprehensive study of major aquifers in Africa published in January 2022 by Scanlon, et al.

It would be good to enter the geographic coordinates along the frame of Figure 1. Nothing else to report regarding the Section “Study area and background”.

We have modified most of the map-based figures as suggested. We elected not to modify Figure 9 because the maps are small in that figure and the scale has been established in the prior figures for the same region.

What uncertainties can be considered for the imputed GWSa data in Figure 7? Can the authors improve the text related to this aspect?

Figure 7 was presented in the Methods section in the context of gap filling. A full presentation of the groundwater storage estimates was provided later in the Results section in Figures 10-13. We added a new section (4.1) to address uncertainty, and included a discussion of “leakage” which was missing from our original text, but is an important element of uncertainty. We believe this strengthens the manuscript and wish to thank the reviewers for this suggestion.

To estimate the Annual Recharge, the authors modify the method of Wu et alii (2019) replacing well measurements with previous computed groundwater storage anomalies (lines 388-390). This is a crucial point, replacing measured values with model-derived values. Do the authors believe that the lack of direct measures may lead to errors in the assessments that are not acceptable for management?

In the original manuscript, the following statement was at lines 388-390:

“For this study, rather than well measurements, we used the computed GWSa for the aquifer and assumed that ∆GWSa = S_y∆h [61].”

However, in Section 1.4 of the introduction we present a background of the Water Table Flucutation (WTF) method and refer to a number of studies including, but not limited to, the Wu et al study where other researchers apply the WTF method to GRACE-derived groundwater fluctuations rather than in situ water level measurements, so we are not “modifying” the WTF method nor are we the first to use it in this fashion. Furthermore, we simply do not have enough historical water levels measurements in the study areas to reliably apply the WTF method with in situ data, as we describe at the end of the first paragraph in Section 1.4:

“Furthermore, these methods rely on in situ measurements and laboratory analysis, both of which are scarce in West Africa.”

It would improve the readability of the manuscript to put the list of abbreviations (GWSa, WFT, CMAP, CANa, and so on) at the end (see template format, Instruction for Authors).

We added a table of abbreviations to the end of the document.

Author Response File: Author Response.pdf

Reviewer 2 Report

1. The manuscript lacks a (sub)section on study limitations. It is advisable to present them in a separate (sub)section.
2. There is Discussion in the manuscript. Is it part of the Results section?
3. I don't quite understand the reasons for the increase in groundwater storage over the last 20 years (Figures 14 and 15). You wrote that "The calculated correlation coefficients are 0.54 and 0.43 in the Iullemeden and Chad basins respectively indicating a good correlation between the two variables." (Lines 467-469). This means that the contribution (as R^2) of precipitation to the long-term variability of groundwater storage is 29% and 18%, respectively. In this case, what are the other (primary) causes that determine the remaining approximately 71-82% of the variability of groundwater storage? This is an important question, since the credibility of the results obtained can significantly depend on the answer to it. Moreover, why is the correlation made only with precipitation, but not with the moisture coefficient, which takes into account not only precipitation but also temperature?

In addition:

1. All maps in the manuscript lack a linear scale.
2. In Figures 2,3, and 4, there are no symbols for different-colored fields and lines.
3. In Figures 5,7,8,10,12,14-16, the variable "groundwater storage" is measured in centimeters on the vertical axes of the graphs. Why? Isn't this a volume variable that should be measured in cubic units of water?
4. In Figure 6, what are the units of measurement on the vertical scales of the graphs?
5. Why did you use the same words and phrases in KEYWORDS as in the title of the manuscript? This reduces the efficiency of their use.

Author Response

Reviewer 2

1. The manuscript lacks a (sub)section on study limitations. It is advisable to present them in a separate (sub)section.

We have integrated limitations into the text of the document. In general, most of the techniques we use have been presented and published in other contexts. These documents provide details on limitations. This case study is the develop and present data on the specific aquifers in Niger, West Africa.

We have made edits in response to this, and other, comments that better explain some of the study limitations.

1. There is Discussion in the manuscript. Is it part of the Results section?

We have integrated our discussion and results section in this manuscript.

1. I don’t quite understand the reasons for the increase in groundwater storage over the last 20 years (Figures 14 and 15). You wrote that “The calculated correlation coefficients are 0.54 and 0.43 in the Iullemeden and Chad basins respectively indicating a good correlation between the two variables.” (Lines 467-469). This means that the contribution (as R^2) of precipitation to the long-term variability of groundwater storage is 29% and 18%, respectively. In this case, what are the other (primary) causes that determine the remaining approximately 71-82% of the variability of groundwater storage? This is an important question, since the credibility of the results obtained can significantly depend on the answer to it. Moreover, why is the correlation made only with precipitation, but not with the moisture coefficient, which takes into account not only precipitation but also temperature?

We thank the reviewer for the detailed comment and analysis. We did not attempt to determine the causes or contributions to groundwater, as we do not have the data to attribute the changes we observe to specific processes. We did show a minimal correlation with long-term precipitation trends, but do not attempt to attribute changes to precipitation, water use, or land cover change. That analysis is beyond the scope of this paper.

In addition:

1. All maps in the manuscript lack a linear scale.

We have modified most of the map-based figures as suggested. We elected not to modify Figure 9 because the maps are small in that figure and the scale has been established in the prior figures for the same region.

1. In Figures 2,3, and 4, there are no symbols for different-colored fields and lines.

We added a legend for these figures.

1. In Figures 5,7,8,10,12,14-16, the variable "groundwater storage" is measured in centimeters on the vertical axes of the graphs. Why? Isn't this a volume variable that should be measured in cubic units of water?

For the GRACE mission it has always been customary to present water storage anomalies in terms of liquid water equivalent in units of centimeters. We are following a long-established practice. Cubic units requires an accompanying area, by use cm of water, trends in various locations can be compared without defining specific areas.

1. In Figure 6, what are the units of measurement on the vertical scales of the graphs?

This has been fixed on all graphs.

1. Why did you use the same words and phrases in KEYWORDS as in the title of the manuscript? This reduces the efficiency of their use.

We have edited and added to the KEYWORDS so they include other topics besides those in the title, though there is still some redundancy.

Author Response File: Author Response.pdf

Reviewer 3 Report

I have reviewed an article on Evaluating Groundwater Storage Change and Recharge using GRACE Data: A Case Study of Aquifers in Niger, West Africa. The manuscript is well discussed. I have the following comments on the structure of the draft.

1. The current introduction has a data description and discussion within it. Details about GRACE can be given separately in a data section. To me introduction has too many sections, it should go in a flow.
2. A little description of the study area should come in your introduction as well. It is all the more important in the case of GRACE as the spatial extent of the study area and its terrain is important to know the usability of GRACE. You can keep the detailed study area section as it is. But the reader need not wonder how big is your area, if it is suitable for GRACE study, etc.
3. At the end of the Introduction, giving a little detail of the structure of your work is a good practice, so that reader gets an orientation on what to expect in the coming paragraphs.
4. In all of your figures, there is no scale bar and coordinates, I cannot make it out how big is your study area. Please add that.
5. GRACE data gap filling plots quality needs to be improved.
6. Figure 9 is GWSa or GW trend? If you mean Anomaly, then of which month?
7. In the data section, you must describe the precipitation data, what spatial and temporal resolution. What is the original source of it, etc.

Author Response

I have reviewed an article on Evaluating Groundwater Storage Change and Recharge using GRACE Data: A Case Study of Aquifers in Niger, West Africa. The manuscript is well discussed. I have the following comments on the structure of the draft.

• The current introduction has a data description and discussion within it. Details about GRACE can be given separately in a data section. To me introduction has too many sections, it should go in a flow.

We acknowledge the comment but prefer to keep the section headings. This allows readers to find sections within their interest and skim those with which they have less interest.

• A little description of the study area should come in your introduction as well. It is all the more important in the case of GRACE as the spatial extent of the study area and its terrain is important to know the usability of GRACE. You can keep the detailed study area section as it is. But the reader need not wonder how big is your area, if it is suitable for GRACE study, etc.

We added a short section to the introduction with general aquifer areas and details in Section 2.2, which describes the study area.

• At the end of the Introduction, giving a little detail of the structure of your work is a good practice, so that reader gets an orientation on what to expect in the coming paragraphs.

We added a few sentences to the end of Section 1.1 to provide an overview of the study.

• In all of your figures, there is no scale bar and coordinates, I cannot make it out how big is your study area. Please add that.

We have modified most of the map-based figures as suggested. We elected not to modify Figure 9 because the maps are small in that figure and the scale has been established in the prior figures for the same region.

• GRACE data gap filling plots quality needs to be improved.

The plots used in this figure have all been updated/improved.

• Figure 9 is GWSa or GW trend? If you mean Anomaly, then of which month?

We added the date to the figure caption. The caption indicates that the figures represent the anomaly in April 2021.

• In the data section, you must describe the precipitation data, what spatial and temporal resolution. What is the original source of it, etc.

We added a description of the precipitation data was added in section 4.3; we used monthly data on a 0.05°×0.05° degree grid

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Thank you for revising the manuscript.