Inferring soil salinity in a drip irrigation system from multi-configuration EMI measurements using adaptive Markov chain Monte Carlo

Jadoon, Khan Zaib; Altaf, Muhammad Umer; McCabe, Matthew Francis; Hoteit, Ibrahim; Muhammad, Nisar; Moghadas, Davood; Weihermüller, Lutz

doi:https://doi.org/10.5194/hess-21-5375-2017

Articles | Volume 21, issue 10

https://doi.org/10.5194/hess-21-5375-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/hess-21-5375-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 21, issue 10

Research article

|

26 Oct 2017

Research article |

| 26 Oct 2017

Inferring soil salinity in a drip irrigation system from multi-configuration EMI measurements using adaptive Markov chain Monte Carlo

Khan Zaib Jadoon, Muhammad Umer Altaf, Matthew Francis McCabe, Ibrahim Hoteit, Nisar Muhammad, Davood Moghadas, and Lutz Weihermüller

Download

Final revised paper (published on 26 Oct 2017)
Preprint (discussion started on 08 Aug 2016)

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'comments on hess-2016-299', Anonymous Referee #1, 07 Sep 2016
- AC1: 'Reply to reviewer comment', Khan Z. Jadoon, 21 Nov 2016

Peer-review completion

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

ED: Publish subject to revisions (further review by Editor and Referees) (19 Dec 2016) by Monica Riva

AR by Khan Zaib Jadoon on behalf of the Authors (26 Jan 2017) Author's response Manuscript

ED: Publish subject to revisions (further review by Editor and Referees) (08 Feb 2017) by Monica Riva

The manuscript have been revised by one reviewer. In their replies, the authors seems confident to be able to address the main issues raised during the reviewer process.
After my own re-reading of the manuscript, I also have the following comments
- In the manuscript (abstract and introduction) the authors talk about parameters uncertainty and model uncertainty. Indeed in their application they only consider “parameter uncertainty” since only one model is considered. This point should be clarified in the manuscript. For example in the abstract the sentence “The model parameters and uncertainty “ is misleading. I guess the authors mean “uncertainty in model parameters”.
- Introduction: page 2 last sentence “Generally ….analysis of parameter uncertainty and correlation is often left unaddressed”. This statement is false. In the literature (and in particular in hydrology – hydrogeology) there is a large number of studies dealing with parameters uncertainty (starting at least from the ’80).
- Page 3, line 7 from the bottom “… to solve the full solution…”please rephrase
- Page 4, eq (1). Symbol x not defined.
- Page 4 eq (4). Syobol j not defined
- Page 4, line 6 from the bottom “….the prior distributions of a given model….”. This sentence is not clear. Do you the authors mean the prior distributions of parameters a given model?
- Page 5, symbol y two lines after eq (6) should not be in bold
- Eq (11). I am confused by this equation for the pdf of sigma. Sigma is defined as a positive (unbounded) variable. Therefore the integral of (11) between 0 and Infinity must be equal to 1. This is not true according to (11).
- Page 6. The authors introduce the parameters to be estimated. These include (in the 3 layer system) 3 conductivities and only 2 layer thickness. Why the thickness of the third layer is not estimated?
- Page 8. Parameters sigma and sigma_b are not defined
- Page 8, line 8 from the bottom. Typo: replace siline by saline
- Please consider to use the help of an expert in English usage and grammar to revise the manuscript

If the authors decide to revise and resubmit the manuscript to HESS they should carefully address these issues together with the main criticisms pointed out by the reviewer.

Hide

AR by Khan Zaib Jadoon on behalf of the Authors (15 May 2017) Manuscript

ED: Referee Nomination & Report Request started (06 Jun 2017) by Monica Riva

RR by Anonymous Referee #1 (20 Jun 2017)

Suggestions for revision or reasons for rejection

The revision of hess-2016-299 is a good improvement compared to the original submission.
The materials and methods are now clear. Conclusions are adequate and supported by the results.
All responses to my comments and to the Editors’s comments are adequate.
I have just one note. In my original report I commented:
1.9: L239 here’s my only methodological issue with this paper. You use as nonconductive scenario a soil that is completely dry. (By the way, what are the salinity values measured at this site? E.g., the conductivity of the saturated paste extract?). In their protocols for use of apparent electrical conductivity measurements in agriculture, Corwin and Lesch state that the soil volumetric water content should be at least 50% of the value at field capacity (ideally between 70% and field capacity. Otherwise, the liquid pathways of electrical conductivity through the soils would be interrupted, unpredictably increasing the resistivity of the soil. This is very likely reason why your results on the non-conductive scenario are not encouraging. My criticism is the following: with one scenario where ECa is known not to be reliable, is the other scenario (highly conductive medium) enough to provide context to your data analyses? I fear not. I think this paper would make much better of a point if other scenarios (e.g., increasing water contents?) were presented. See: Corwin, D.L., and S.M. Lesch. 2013. Protocols and guidelines for field-scale measurement of soil salinity distribution with ECa-directed soil sampling. J. Environ. Eng. Geophysics 18(1):1-25. and: Corwin, D.L., and S.M. Lesch. 2005b. Characterizing soil spatial variability with apparent soil electrical conductivity: I. Survey protocols. Comput. Electron. Agric. 46(1-3):103-134.
Authors responded:
Reply 1.9: We thank the reviewer for highlighting this important issue. For the same site, Jadoon et al. (2015) reported a relationship to relate bulk electrical conductivity to the soil salinity (i.e., the conductivity of the saturated paste extract). Observed soil salinity range between 3-185 dS/m). As discussed earlier (Reply 1.1), in the revised manuscript same relationship will be used to estimate the soil salinity. Text and Figure C5 HESSD Interactive comment Printer-friendly version Discussion paper 9 will be incorporated to show the soil salinity distribution. Synthetic and field measurements were analysed to test the performance of the electromagnetic forward model in conductive and non-conductive soil, and retrieve soil salinity using Bayesian inversion. In the case of synthetic scenarios, EMI data was generated using electromagnetic forward model and Bayesian inversion was used to estimate five parameters (three layer electrical conductivities and two layer thicknesses). Result shows that the electromagnetic forward model is not sensitive to the non-conductive soil. Similarly, Minsley (2011) used synthetic data considering the characteristics of shallow ground-based EMI system, geophex GEM-2 and reported that the electromagnetic forward model is less sensitive to the non-conductive soil. Indeed, in the agriculture field the soil electrical conductivity decreases if the soil water content is below 50% of the field capacity, which may cause the less encouraging results for the nonconductive soils. This issue will be highlighted in the manuscript and the references of Corwin, D.L., and S.M. Lesch. 2005 and 2013 will be incorporate.

While I am satisfied by this reply, I think I might have missed where this issue was addressed in the revised manuscript.

Congrats for the very interesting research!

Hide

ED: Publish subject to minor revisions (further review by Editor) (06 Jul 2017) by Monica Riva

AR by Khan Zaib Jadoon on behalf of the Authors (16 Jul 2017) Author's response Manuscript

ED: Publish as is (23 Jul 2017) by Monica Riva

AR by Khan Zaib Jadoon on behalf of the Authors (09 Aug 2017)

Short summary

In this study electromagnetic induction (EMI) measurements were used to estimate soil salinity in an agriculture field irrigated with a drip irrigation system. Electromagnetic model parameters and uncertainty were estimated using adaptive Bayesian Markov chain Monte Carlo (MCMC). Application of the MCMC-based inversion to the synthetic and field measurements demonstrates that the parameters of the model can be well estimated for the saline soil as compared to the non-saline soil.