Investigating the environmental response to water harvesting structures: a field study in Tanzania

Eisma, Jessica A.; Merwade, Venkatesh M.

doi:https://doi.org/10.5194/hess-24-1891-2020

Articles | Volume 24, issue 4

https://doi.org/10.5194/hess-24-1891-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/hess-24-1891-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 24, issue 4

Research article

|

16 Apr 2020

Research article |

| 16 Apr 2020

Investigating the environmental response to water harvesting structures: a field study in Tanzania

Jessica A. Eisma and Venkatesh M. Merwade

Download

Final revised paper (published on 16 Apr 2020)
Preprint (discussion started on 26 Apr 2019)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

SC1: 'Cooment on 50% of sand dams functiong', Giulio Castelli, 06 May 2019
- AC1: 'Reply to SC1: Comment on 50% of sand dams functioning', Jessica Eisma, 06 May 2019
RC1: 'Review of “Investigating the environmental response to water harvesting structure: a field study in Tanzania”', Rolf Hut, 29 May 2019
- AC2: 'Reply to RC1', Jessica Eisma, 27 Jun 2019
RC2: 'Review', Anonymous Referee #2, 24 Jun 2019
- AC3: 'Reply to RC2', Jessica Eisma, 30 Jun 2019
AC4: 'Summary of Responses to RC1 and RC2', Jessica Eisma, 08 Jul 2019
RC3: 'Investigating the environmental response to water harvesting structures: A field study in Tanzania', Anonymous Referee #3, 12 Jul 2019
- AC5: 'Response to RC3', Jessica Eisma, 12 Jul 2019

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (further review by editor and referees) (22 Jul 2019) by Graham Jewitt

AR by Jessica Eisma on behalf of the Authors (28 Aug 2019) Author's response Manuscript

ED: Referee Nomination & Report Request started (17 Sep 2019) by Graham Jewitt

RR by Rolf Hut (19 Sep 2019)

RR by Alison Parker (02 Oct 2019)

RR by Tibor Stigter (13 Dec 2019)

Suggestions for revision or reasons for rejection

The manuscript covers a nice study on the environmental response to sand dams and does so in a fairly complete way, by looking at water levels, erosion, vegetation growth and macroinvertebrate community behaviour. My main remark is that the collected groundwater level data allow a more comprehensive overview of the water table response at the three sand dam sites, including the non-functioning one. A total of 63 wells were drilled and water levels were monitored over one year (most successfully at Chididimo), which gives much more insight into the water table dynamics (in time and space) than described and discussed in the text. Moreover, the borehole logs from the 63 wells provide valuable lithological information for understanding connectivity between the (sand) river and its banks, which is now poorly addressed. The other component that I find a bit missing is the water quality aspect, which links directly to the apparently dominant mechanism of water loss in the areas, namely evapotranspiration (ET). If ET is indeed that dominant, as compared to lateral or vertical seepage, this should be seen in the increasing salinity of the water throughout the dry season, especially after a few years (the Soweto sand dam for instance was completed in 2011). It would be interesting to see if there is any additional information from EC measurements or interviews with the community that evidence a problem of increasing salinity.

I have further detailed my comments and suggestions for revision for each section below.

Abstract: Is well-written, but will need some editing after including additional results and discussion. It would be useful to read here why sand dams are not a suitable habitat for macroinvertebrates.

Introduction: sufficiently comprehensive and to the point, with two exceptions: the response of the groundwater table in the surrounding area of a sand dam largely depends on the connectivity between that river margins and the sand river. This connectivity is often seen to be limited, which in part explains why the effect of an increasing water table is very local. This is not addressed in the text. The other aspect I would like to see introduced is that of sand dam water quality and processes (and/or human activities) that affect it.

Study area:
- In the description of climate I would like to see average annual values of P, T and PET.
- The description of the geology is lacking. This is crucial information to discuss the connectivity between the (sand) river and its banks. As soil logs were collected from each of the 63 drillings, there is a lot of information available.
- The characterization of the dimensions of the sand reservoirs built up behind the sand dams is missing. It would be important to state something about thickness, length and estimated storage capacity. In combination with the extinction depth this reveals the depth of “safe storage” not affected by ET. Only the maximum width (at the dam site) has been indicated for each area.
- An overall schematic diagram of a sand dam, and how it matures over time, would also be very useful. This would further help explain the non-functionality of the Kimokouwa sand dam.
- The location of the hand pumps is not clear for the Kimokouwa and Soweto sand dams (where is “near the sand dam”?). For Chididimo why was the hand pump installed so far away from the dam? 180 m is beyond the reach of influence of the dam, meaning that at that location there is hardly and noticeable increase in sand and water storage.
- Why was there no observation well placed directly behind (upstream of) the sand dam, in the middle of the river? This would have allowed to measure the fluctuations that are not affected by transpiration (only evaporation, seepage and abstraction).
- For the water table monitoring wells (WTMW) was topographic elevation of the top of the pipe measured, and with what accuracy?
- The water balance and resulting volumetric water loss calculations require some clarification. It is not clear how the “control volume” and corresponding water volume were calculated. I understand from the text that water volumes were integrated over the area covered by the monitoring wells. But how did the authors take the spatial variation in texture and related porosity into account? How did the texture of the margins compare with that of the river bed? What was the spatial distribution of the water table and the spatial variation in water table response to the sand river. Did the water table in the margins actually respond to the river level recharge or to aerial recharge from rainfall? This could be shown through piezometric maps at different moments of the year (end of the wet season and dry season) to check the losing or gaining behaviour of the sand dam. With so many well level recordings in time and space this invaluable information could be provided. It will also allow showing the importance of ET vs. direct soil evaporation (at places and/or times in the year where/when vegetation is absent). It will further allow to assess if lateral flow towards the river margin occurs, as hypothesized by the authors.
- FLDAS calculations should be better explained. In section 4 the authors mention that the FLDAS dataset calculates evaporation from bare soil based on simulated soil moisture content. How does this compare to transpiration, and how is the reduction of E and T with depth taken into account?
- Canopy-interception of rainfall should be entirely excluded from the water loss calculations, unless you include rainfall as an input to the water balance. If only ET from the subsurface is considered (water table, soil moisture) then including canopy interception makes no sense, even in the riparian corridors.
- Please explain how Qsb(t) (baseflow-groundwater runoff) was determined in the field. It is not mentioned.

4. Sand dam impact: I recommend to name this section Results and Discussion, which you then separate per topic. Section 5 could then be General Discussion and Considerations, and then the Conclusions section is perhaps not really necessary.

4.1 Macroinvertebrates: clear and interesting.

4.2 Vegetation: The authors provide evidence to support the statement that “it is reasonable to expect that the vegetative cover at the Soweto and Chididimo sand dams is improved, in part, by a locally raised water table near the ground surface”. The arguments are clear, but could be further supported by the water level measurements, which are not shown. I do wonder why at the lowest elevation vegetation is most sparse in the dry season. Moreover, I do not agree that the fact that Kimokouwa has a very low vegetation cover in both the dry and wet season is related to a poorly functioning sand dam. It seems much more related to the topography, with higher elevations and slopes, and hence higher runoff and lower infiltration potential. For vegetation the role of groundwater can be important, but the role of the infiltration and water-holding capacity of the soil can be at least as relevant. In finer soils and flatter topography you will therefore expect more vegetation. The role of the sand river is questionable, with a few exceptions shown in Figure 3 (VT4 wet season Soweto, VT2 dry season Chididimo, and perhaps VT3-4 wet season Kimokouwa).

4.3 Streambank erosion: clear and well supported by the observations; I am only not sure why the downstream curve for Chididimo is missing.

4.4 Water Table: As mentioned, this section could be enriched by showing the results of the 12-hourly measurements taken at Chididimo and Soweto, to show the temporal and spatial variations of the water table and how they related to the sand dam water level. The borehole log data could be shown to address the connectivity between river and banks, and the potential for groundwater flow into the banks. In many cases this connectivity is actually rather limited, even when sand dam water level is able to rise above the regional water level in the margins. That is probably another reason why the sand dam at Kimokouwa was non-functional, as the river valley is steeply incised (as in figure 5a), creating a small narrow sand reservoir with limited storage capacity and no possibilities to actually feed the neighbouring water table, which is expected to be above the sand dam level. Again, this could be verified with the water level data where they are available.
- How is the 1% of the total water lost attributed to baseflow determined?
- In Figure 8 show how the red line was calculated and add the ET line for comparison.
- How does ET vary spatially in the areas? This is shown by the water table fluctuations? How do the variations below vegetative cover (dominated by ET) compare to those below bare soil (in particular the sand reservoir area, dominated by E)?
- The fact that Chididimo and Soweto experience approximately 100 mm lower rainfall than in Kenya is not the cause of faster storage depletion throughout the dry season. It is the length of the rain season and the duration/frequency of river flow that determines whether a sand dam lasts longer. This is actually addressed in the discussion.
- The authors state correctly that “The Soweto sand dam is losing water during the shallow ET phase at nearly 2.5 times the rate of the Chididimo sand dam.” They argue that this is because: “The width of the Soweto sand dam is nearly twice that of Chididimo, providing a greater surface area of sand from which evaporation occurs”. However, the rate is already given in mm/day, independent of area. It is more likely, also by looking at the elevations in Figures 1c/1d, that the water levels are overall much shallower in the Soweto area. This would also show in the water table monitoring data. In that case, the grey area in Figure 8a would actually correspond to “Deep ET” or “Intermediate ET” in Chididimo, rather than shallow ET. The difficulty here is that for the calculations not only the sand dam itself is considered, but also the surrounding area. It would be interesting to compare this to the calculations only for the sand reservoir behind the sand dam.

5. Discussion
The authors mention that “the functioning sand dams had significant impact on the local water table”, but do not provide the evidence to support this. This would require knowledge on the water table dynamics before the sand dams were constructed, or the thickness of the sand deposits/total capacity of the sand reservoir before the sand dam was built. It is obvious that there will have been impact, but at least try to show it through the observed water level behaviour after the dam was built, showing the extent to which the sand dam had an effect.

Hide

ED: Publish subject to revisions (further review by editor and referees) (23 Dec 2019) by Graham Jewitt

AR by Jessica Eisma on behalf of the Authors (03 Feb 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (04 Feb 2020) by Graham Jewitt

RR by Tibor Stigter (10 Mar 2020)

ED: Publish as is (18 Mar 2020) by Graham Jewitt

AR by Jessica Eisma on behalf of the Authors (18 Mar 2020) Manuscript

Short summary

Sand dams capture and store water for use during the dry season in rural communities. A year long field study of three sand dams in Tanzania showed that sand dams are not a suitable habitat for aquatic insects. They capture plenty of water, but most is evaporated during the first few months of the dry season. Sand dams positively impact vegetation and minimally impact erosion. Community water security can be increased by sand dams, but site characteristics and construction are important factors.