Understanding runoff processes in a semi-arid environment through isotope and hydrochemical hydrograph separations

Camacho Suarez, V. V.; Saraiva Okello, A. M. L.; Wenninger, J. W.; Uhlenbrook, S.

doi:https://doi.org/10.5194/hess-19-4183-2015

Articles | Volume 19, issue 10

https://doi.org/10.5194/hess-19-4183-2015

© Author(s) 2015. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/hess-19-4183-2015

© Author(s) 2015. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 19, issue 10

Research article

|

20 Oct 2015

Research article |

| 20 Oct 2015

Understanding runoff processes in a semi-arid environment through isotope and hydrochemical hydrograph separations

V. V. Camacho Suarez, A. M. L. Saraiva Okello, J. W. Wenninger, and S. Uhlenbrook

Download

Final revised paper (published on 20 Oct 2015)
Preprint (discussion started on 22 Jan 2015)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC C188: 'Review of the paper entitled "Understanding runoff processes in a semiarid environment through isotope and hydrochemical hydropgraph separation"', Anonymous Referee #1, 16 Feb 2015
- AC C1295: 'Answer to Referee comments', Vivian Camacho, 01 May 2015
RC C217: 'Referee report on the manuscript "Understanding Runoff Processes in a Semi-Arid Environment through Isotope and Hydrochemical Hydrograph "', Anonymous Referee #2, 18 Feb 2015
- AC C1296: 'Answer to Referee comments', Vivian Camacho, 01 May 2015
AC C1294: 'Authors' response to Referee1 and Referee2', Vivian Camacho, 01 May 2015

Peer-review completion

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

ED: Reconsider after major revisions (24 May 2015) by Axel Bronstert

AR by Vivian Camacho Suarez on behalf of the Authors (04 Jul 2015) Author's response Manuscript

ED: Referee Nomination & Report Request started (19 Jul 2015) by Axel Bronstert

RR by Anonymous Referee #2 (03 Aug 2015)

Suggestions for revision or reasons for rejection

I found that many concerns of the other reviewers have been addressed and also some of my minor concerns, which I acknowledge. However, my major concerns have not been handled adequately. They are very basic and call all the main findings and the main conclusions of the paper into question.

I restate my concerns down below:

1.) I still doubt that the presented case study investigates a truly dry and semi-arid system. Normally, this distinction is not that important but in the present case it is: The study aims at transferring and testing tracer techniques from humid catchments (here they have been widely used) to the “semi-arid” Kaap catchment. If the Kaap is not typically semi-arid but more or less humid, the obtained results and conclusions are nothing new but just another case study and not transferable to really semi-arid systems, where other researchers might plan similar investigations. For those investigations the conclusions of the present paper will be misleading and lead to wrong expectations concerning the benefit of tracer and isotope-based hydrograph separations. I therefore repeat my concerns according to the character of the Kaap catchment.

a.) In the discussion paper the original value for PET “from Class A evaporation pan from SASRI and DWA stations (2002 to 2012)” was 1014 mm (line 23, p 980, discussion paper). Obviously, these stations are still used to calculate actual evaporation (P.6, line 15-24). Why aren’t these stations used anymore to characterize the climate of the catchment? Instead the authors present a map (Figure 2F) with PET-values almost double (1663 – 1998 mm) on which they base their climate index map on (Figure 2E). While annual rainfall may vary considerably, it is hard for me to imagine that a ten year PET-average from several stations shows such large deviations from long term data. How can this be explained? To me the newly produced index map seems to exaggerate existing Class-A-Pan data from within the catchment, or simply uses them without correction.

b.) The authors state that Köppen-Geiger classifies the Kaap as “sub-tropical”. This is wrong, because “sub-tropical” is a geographical term to describe the location, and used in terms of climate it only includes temperature but not precipitation. Köppen-Geiger includes both precipitation and temperature. According to Köppen-Geiger, the Kaap is in C (temperate) and not in B (arid). By the way, why do the authors produce a new map and do not use existing ones? Also according to the well-known map of Meigs (1953), the Kaap is not semi-arid.

c.) All these classifications are based on annual index values. To really characterize a hydrological system, seasonality plays an important role. The present study intends to describe hydrology only using four events in one wet season. If an annual indicator is in the boarder between dry and wet, it will truly be wet when calculated for the wet season or for the month in which the investigations were carried out. There are monthly indicators around: Use them for the season you carried out your investigations and you may describe the actual aridity of your system at the time you took the measurements!

d.) This brings me to another point: hydrology. I am pretty sure that hydrologically the catchment may be treated as a humid system (at least during the wet season) and all hydrological indicators provide evidence for this. This I pointed out in my first review, but the authors totally ignored this point in their response to the reviewers and in their revised version. Therefore, I quote word by word the relevant passage of my first review on hydrological characteristics that was left unanswered by the authors:
“Although the minimum daily flow is 0, there is still a monthly average of 0.8 m3s-1 after the dry season (p980). This indicates that periods of zero streamflow are very short and that your channels are not ephemeral which they should be, if you are in a really dry region. Another indication for a humid system is the fact that you have a gaining river system (p 987). This means there are constant flow paths through the subsurface, and constant baseflow (see above). Then a relevance of groundwater is logic also during runoff events, because piston-flow mechanisms may activate existing subsurface flow paths. This is in accordance with your tracer results that show the dominance of pre-event water during runoff events. You state yourself that this well known for “semi-humid” regions (p977). This means that these results confirm existing knowledge from humid systems but do not contribute to process knowledge for semiarid environments, which you aim at”

2.) Also my second major concern was not adequately addressed. While the runoff component definitions are a little clearer now, there is still the component “direct runoff” which is separated from groundwater components. “Direct runoff” or “quick flow” is referred to as direct precipitation into the river (this can be neglected due to the small area of open water in dry regions) and to Hortonian overland flow (p 13). Even if cited in a 15 year old paper by actually the same leading author, this definition is problematic as it mixes dynamics and source areas. One must differentiate between separations based on dynamics (direct runoff/quick flow in contrast to base flow) and separations based on source areas (groundwater in contrast to surface components). If one looks at the small percentages of direct/quick flow during all studied events - and this is true for both the two- (figure 5) and the three-component (figure 9) hydrograph separations - this error becomes obvious. In hydrological textbooks quick flow (direct runoff) is usually defined as the fraction of flow that reacts to rainfall and is distinguished from the slow flow (base flow) component. How can you explain a sharp runoff response almost without “quick flow”?

3.) And finally, a third very basic concern remains: the blurred definition of the component “shallow groundwater”. This component is postulated to be the “intermediate component from perched groundwater tables” (p 9) or the “near-surface, soil water component” (abstract). It is only characterized by “initial stream water samples before rainfall started”, but not by a single real water sample from e.g. a groundwater well. Hence its definition is arbitrary and purely speculative. According to the mixing diagram (Figure 8) its main characteristic is elevated potassium. Already in 1996 Sandstrom (HYPROC 10, 703-720) used major ions to study runoff generation. He worked in a real semi-arid area (Tanzania). He directly characterized the source areas by sampling and found out that potassium was elevated already by throughfall. I do not see any justification why in the Kaap potassium should characterize shallow perched groundwater. It may be enriched both by canopy and by soil contact. So why shouldn’t it be elevated also in Hortonian runoff? But Hortonian runoff is just the other end member (“direct/quick flow”).

In summary all conclusions of the paper are problematic because its statements:
1.) “dominating groundwater contributions” are based on blurry and arbitrary end member definitions and
2.) “applicability of tracer based hydrograph separations methodology for semi-arid regions” are based on investigations inside a wet system and by no way are representative for really dry environments.

Hide

RR by Anonymous Referee #1 (03 Aug 2015)

ED: Publish subject to minor revisions (Editor review) (20 Aug 2015) by Axel Bronstert

Dear Authors,

being the handling editor of this manuscript, I now have the task to decide how to proceed with this submission. This is actually not an easy task, having the revised version on the desk and two reviews from well experienced reviewers. Those two reviews come to the maximum possible opposing conclusions: “Reject” (Rev. 1) vs. “accept as it is” (Rev 2).
First, I add the explanations from Rev.#2, which he added as “confidential to the editor” (probably he just did this accidentally).
“I recommend the publication of the paper in its actual form. Despite my concern during the first review about the reduced amount of samples analysed has not been solved (the authors have the samples they have and they cannot have any more) I consider that the authors have done a very profound review following the comments of myself and the second referee and the paper has improved greatly, especially concerning the introduction and discussion sections. For this reason, I think that the paper deserves to be published in HESS as it is.”
Second, I try to understand and assess the rather negative review from Rev # 1:
1) Criticism of the climatic classification of the Kaap Catchment. It is argued that the catchment is not semi-arid but semi-humid. Well, there is no very crisp and unique definition of semi-aridity.
Therefore, I had a look in Schulze et al, 1997, which is an excellent display of South African Agrohydrology and –climatology (though not cited in the paper neither the rview !). This is rather good in accordance with your figure 2 (which I find well placed). From this I can derive that in any month the pot ET is higher than the Precip. Therefore, one may keep the term semi-aridity. Of course it is sub-tropical, but his is not of major relevance. There are many possible peculiarities of semi-arid climate, even if rainfall is as high as 1000 mm /year. However, one can also derive that the small scale spatial variability of rainfall is high in this region, probably due to the distinct topography of this are.
2) Questioning if the hydrology should be treated “as a humid system (at least during the wet season)”. Well, again, there is no overall accepted standard (and no need) for a clear and crisp differentiation between humid and arid hydrological systems. Actually, many (but not all) hydrological systems have both dry and humid characteristics. Therefore, I do not see why this study should be not considered as a study in a semi-arid environment. I have been participating in semi-arid research myself, with conditions of rainfall as high as 800mm and pET above 2500mm. Groundwater contribution to streamflow (more precise: two-way interactions of alluvial groundwater and channel water) was an important mechanism for streamflow generation, of course during the wet season and at the beginning of the dry season. I definitely agree with Rev#1 that the processes an (therefore the validity of the results) do differ between the dry and the wet season
3) Questioning if the relatively sharp increase in runoff (“quick flow”) can be triggered at least partly by groundwater exfiltration. In case there is a hydraulic shortcut (preferential flow) between alluvial groundwater / shallow groundwater and channel system (and infiltrating rainwater and the shallow groundwater) this is a possible mechanism. There is ample of literature who reports on such specific investigations.
4) “blurred definition of the component ‘shallow groundwater’”. I agree, if groundwater is considered to be a major contributing source to runoff. It would be good to describe the groundwater system more in detail.
Therefore, taking into account BOTH recommendations from Rev.#1 and rev.#2 and my own inspections, my conclusions to the authors are: Please prepare a 2nd revised version of this manuscript, where you
• elaborate more on the climate of this catchment. Can you show or produce (from your data) a kind of Walther-Lieth diagram for one or two climate stations in that area?
• Discuss more on the general (semi-arid?) hydrology of this area. Can you show a “typical” hydrograph of the KAAP river (mean runoff, variations)? How many days without runoff/year? Include a new paragraph on the value of groundwater contributions for runoff
• Do you have any indications for quickflow contributions via the groundwater (i.e. indications of hydraulic shortcuts from infiltration to the GW-surface and further ahead towards the Channel?
Mentioned References:
- Costa, A.C., Foerster, S., de Araújo, JC., Bronstert, A. (2013): Analysis of channel transmission losses in a dryland river reach in northeastern Brazil using stream flow series, groundwater level series and multi-temporal satellite data. Hydrological Processes, 27, 1046-1060.
- Schulze RE et al. (1997) South African atlas of agrohydrology and climatology. Water Research Commission, SA

Best regards, A. Bronstert (handling editor)

Hide

AR by Vivian Camacho Suarez on behalf of the Authors (30 Aug 2015) Author's response Manuscript

ED: Publish subject to minor revisions (Editor review) (04 Sep 2015) by Axel Bronstert

Dear colleagues,

thank you for your moderate revision of the paper. However, there are still some open questions which must be dealt with:
1. First open question concerns mainly to your conclusion that you detected a "shallow groundwater" contribution in your hydrograph analysis. You derive this conclusion on the increased potassium concentrations. Thus, the question is: how can you discriminate between surface runoff and shallow groundwater runoff. Potassium can be termed a trace for quickflow, including surface flow.
2. It would help a lot if you can add some more explanations on the groundwater situation there, both deep (or “normal") groundwater and shallow (alluvial, perched?) groundwater. Do you have any data from gw-observation wells? If yes, please include such data in the paper. If no, you need to consider the possibility that the K-concentrations may be due to surface runoff.
3. Finally, If groundwater contributions (both deep and shallow) are that important, it would help the better understanding of the regional conditions, if you discuss and discriminate the different processes during the wet and dry periods of the annuals cycle. If I understood your study correctly, your tracing was performed (only ?) in wet season events. E.g. these processes would be rather similar to the ones in humid regions (as rev. # 2 pointed in his comments). This would require that you make this clear in the conclusions (e.g. you analyse rainy season hydrology in a semi-arid climate, etc.) and that you may discuss (or speculate?) what processes might be of importance for dry runoff generation. Obviously there is runoff in the dry seasons, even if this is small.

I hope that this ongoing discussion helps to further improve the paper and remain with best regards,

Axel Bronstert (handling editor of this manuscript)

Hide

AR by Vivian Camacho Suarez on behalf of the Authors (02 Oct 2015) Author's response Manuscript

ED: Publish as is (05 Oct 2015) by Axel Bronstert

AR by Vivian Camacho Suarez on behalf of the Authors (11 Oct 2015) Manuscript

Short summary

Isotope and hydrochemical tracers are tested providing new insights to isotope hydrograph in semi-arid areas in southern Africa. This study provides a spatial hydrochemical characterization of surface and groundwater sources, end member mixing analysis, and two- and three-component hydrograph separations. Results showed that the Kaap catchment is mainly dominated by groundwater sources, and direct runoff is positively correlated with the Antecedent Precipitation Index during the wet season.