Displacement Analysis of Geothermal Field Based on PSInSAR And SOM Clustering Algorithms A Case Study of Brady Field, Nevada—USA
Round 1
Reviewer 1 Report
Dear Authors, I have just complete the review of your manuscript. The article is well written and it is very intriguing. I did some minor remarks directly on the pdf file that you will find attached to this letter.
Two main special remarks.
1)Please explain why you decide to analyse only a period 2 years of sar images. The study area shows a very slow trend of subsidence (or uplift) ranging in few meter per year if you expand the period of investingation it is possible to obtaine a more consolidated results. Please take into consideration this changing.
2)Try to explain the results witha geologic approach. You stated that subsidence is mostly related to 3 main causes (pore pressure and desaturation, and transportation of the dissolved salt flow ), but i am curios to understand (and it is possible that all the readers would ask the same thing) why the subsidence area and uplift sector are perfectly separated by a NE-SW edge. It is well known that subsidence generates a litle uplift of surroundings but in your case the two opposite movements shows a same range. It seems to a strong control by the structural features. is it possible?
Please add more information about this point
Comments for author File: 
Comments.pdf
Author Response
Please see the attachment.
Author Response File: 
Author Response.docx
Reviewer 2 Report
In my opinion, the innovation of this paper is not sufficient for being published in the RS, and the experiments and results are not convincing.
As mentioned in the paper, the Brady Field had been investigated by the InSAR technique by other authors. Compared to the previous results, the PSI deformation results are quite sparse. Actually, PSI algorithm is not suitable in the study area due to the scarce of PSs. The SOM Clustering algorithm seems to be unnecessary.
In addition, the manuscript is quite preliminary. There are many typing errors which can not be listed one-by-one.
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The paper "Displacement Analysis of Geothermal Field Based on PSInSAR And SOM Clustering Algorithms -- A Case Study of Brady Field, Nevada – USA" by M. Cavur et al. (Manuscript ID: remotesensing-1045160) present a PSI analysis of the Brady Hot Spring Geothermal field using Sentinel 1 SAR data. This paper is one of the similar recent studies of the field referenced in [24] and [53]. In this sense, it contains somewhat incremental science. The main differences are (1) the use of more recent SAR data, between 2017 and 2019, and (2) introducing a method to find the inflection dates possibly caused by geothermal operations. I think the paper is worth publishing but after a substantial revision, especially because I find the novel method (2) poorly presented and the results debatable. Below I list my major concerns, and then offer a list of relatively minor suggestions for further improvements.
Major comments and questions:
(A) In Sect. 3, you mention that the SARPROZ software was used for the PSI analysis. What about the implementation of the SOM algorithm? Did you use an external software product (if yes, please cite it)? Or did you implement the method yourself?
(B) What are the error bars on the derived subsidence/uplift velocities? There is no assessment of measurement errors in the paper at all, and giving the values up to 2 decimal digits (in the text) or even 3 (!) decimal digits (in Fig. 6) suggest that the errors are in the order of ~0.001 mm/yr. This seems overly optimistic. Please estimate the velocity errors and then cite the numerical values only up to the last significant digit.
(C) The displacements are called "subsidence" and "uplift", suggesting that they are true vertical motions. On the other hand, it is well-known that the Sentinel 1 SAR viewing direction is not vertical but inclined. Please make clear in the text if these numbers are to be understood as line-of-sight motions, or perhaps transformed to the vertical. In the latter case, state the assumptions used (i.e. no horizontal motion).
(D) On pages 10-11: Please explain what 3x3 means in "3x3 SOM analysis". Why is it justified to use 9 clusters when some of them obviously show much similar motion characteristics as well as overlapping localization (SOM 5, 6, and 9 in particular; see Fig. 7)? Especially because later the 9 clusters themselves are further manually "clustered" into three types: uplift, subsidence, stable.
(E) In Sect. 4.2: Finding the inflection dates from the 2nd derivative of the displacements is in principle an excellent idea. But how accurately can this be done in the practice? What are the realistic errors in determining the 2nd derivatives displayed in Fig. 9? (By the way, the units are missing from the vertical axis here and in Fig. 10.) Why did you choose 0.25 mm/yr^2 as an "experimental" threshold of reliability? The uncertainties in the values seriously affect the uncertainty of identifying the possible inflection points. Please try to carefully assess the errors associated with this method. Is it physically realistic to find new inflection dates within intervals of a few months, and even within 12 days (Jan and Jun 2019), i.e. at consecutive SAR acquisition epochs?! In my opinion, this section needs a more careful treatment and a clearer/more detailed description of the method in order to prove that the results are solid. It is also unclear to me (and presumably to some potential readers as well) what the orange dots labelled as "delta" or "difference" indicate along the horizontal axes in Figs. 9 and 10.
(F) Related to the previous point, is there any independent way to verify the inflection dates? Do the geothermal activities change over such time scales at all, and how rapidly the surface could react to these changes? I understand that "it needs some field and detailed research that we address for future studies", but at least some theoretical expectations could be discussed here whether the changes in the production are able to cause such inflections at all, if they indeed exist.
Minor comments:
page 1:
The authors' affiliations are completely missing! Instead, positions (Prof., Assistant Prof., etc.) are given which are not needed.
Abstract: explain the abbreviations here as well (SAR, PSI, SOM? etc.)
page 2:
unmanned vehicle -> unmanned aerial vehicle
data and study area [are] explained
page 3:
COPERNICUS -> Copernicus
2.1. Subsection -- give a proper title for the subsection
available every 12 days -- for the 2-satellite system, it is actually 6 days
We downloaded 72 descending 144 orbit SAR images -- What does it mean? Is 144 the track number? Perhaps it is enough to give in the table.
of SAR 1A images -> Sentinel 1A images?
page 5:
3.1. Subsection -- give a proper title for the subsection
page 6:
labels in Fig. 3 are invisibly small
page 8:
labels in Fig. 5 are small
page 9:
When applying Equation (2), the value of n=3 is used, isn't it? It would be appropriate to mention.
and but a different source -> but a different source
page 12:
Table 3., last row: -.0.03 (an extra decimal point)
page 15:
units are missing from Table 5 (presumably mm/yr)
page 18:
Author Contributions:
The text should be cleaned up from the instructions that were originally in the template
Acknowledgements:
"Sentinel satellite missions operated by the German Space Agency (DLR) were used" -- are Sentinel 1 satellites indeed operated by DLR, and not ESA??
Conflicts of Interest:
The text here contains all the examples from the template. Please choose one statement that is appropriate.
page 22:
"Publisher’s Note: MDPI stays..." -- start in a new line
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
The paper had been improved much compared to the first version. Most of my comments had been addressed. But I still have three comments.
- Although the author said the PSI algorithm is used due to its high accuracy, recent years some advanced MT-InSAR methods had been proposed to exploit PS as well as DS. In my opinion, these PS&DS methods are more suitable in the study area. The author should first cite and review these research, and then clarify clearly the reason that they choose the PSI algorithm rather than these PS&DS methods.
- How to determine the thresholds for the subsidence and uplift? Why the subsidence ranges from -19 mm/yr to 1 mm/yr, and uplift range from -2 mm/yr to 14 mm/yr? Please explain the overlap between the subsidence and uplift.
- Figure 3 has very low quality. The legend of right subplot is missing. What do you mean by GCP? It seems that only one GCP is used. Is it the reference point? Which DEM do you used in the differential InSAR?
Author Response
Please see the attachment.
Author Response File: Author Response.docx
Reviewer 3 Report
The manuscript improved substantially after the revision. Now I believe the paper can be published after some minor (but quite numerous) corrections which I detail below.
page 1, Abstract:
applied to the clustered data was to understand -> applied to the clustered data to understand
How can a subsidence be positive (+1 mm/yr) and an uplift negative (-2 mm/yr)? Note that in Sect. 4.1 you state: "The highest subsidence detected in this analysis is -21 mm/yr whereas the highest uplift detected is 14 mm/yr", and the former value seems different from the -19 mm/yr written in the Abstract. Please make the numbers consistent.
on the NE of the field where the zone -> on the NE of the field where there is the zone
"On the other hand..." -- start a new sentence
page 2:
InSAR analyses techniques -> InSAR analysis techniques
page 3:
data is evenly spaced -> data are evenly spaced
page 5:
"we limited our dataset to Sentinel 1A as the Sentinel-1A products are available with a revisit cycle of 6 days for the 2-satellite system, which allowed us to construct time series data for a short time period" -- it is unclear from this sentence if only Sentinel-1A data were used (single satellite = 12-day revisit time) or both 1A and 1B (which together provide the 6-day revisit time); form Table 1, the former case seems to be true. They why do you mention the 2-satellite system and the 6-day revisit?
SAR data, which was used -> SAR data, which were used
Sentinel data for the area is available -> Sentinel data for the area are available
analyses -> analysis (occurs twice)
page 7:
The last sentence is incomplete: "Most of the displacement points in the graphs given in"
page 10:
The SOM analyses -> The SOM analysis
page 17:
the average displacements -6 mm/year -> the average displacement is -6 mm/year
page 18:
it was observed subsidence at -> subsidence was observed at
"13.5 ± 2.9 mm/yr by using GPS data and 9.9 ± 3.3 mm/yr by using InSAR" -- to avoid confusion, here I suggest using the same convention as earlier, i.e. using negative values (-13.5 and -9.9) for subsidence
page 20:
we first find the second derivative -> we first found the second derivative (or: we first calculated the second derivative)
page 22, conclusions:
It is again confusing to call -2 mm/yr as "minimum uplift" while it is presumably a subsidence (well, I would rather call it stability, given that the error bar may be comparable in size)
References:
The list is put together in a rather sloppy way, it needs a very careful check which is not supposed to be the reviewer's task! In general: please check all volumes, numbers: some of them seem missing or incorrect. Also, some titles, journal names are missing or incomplete. Specific mistakes I spotted:
[1] los humeros -> Los Humeros
[11] Bardi et al. -> list all authors
[19] extra "(" to be removed
[20] do not capitalize the title
[25] Ali et al. -> list all authors
[32] Raspini et al. -> list all authors; journal name/volume/paper ID missing
[36] Lazecky et al. -> list all authors; do not capitalize the title
[37] sentinel-1 -> Sentinel-1; insar -> InSAR; bibliographic details are unclear
[44] bibliographic details (publisher) missing
[47] bibliographic details missing
[48] title missing, author affiliation (!) is given instead
[49] title correctly: Terrain Monitoring in China via PS-QPS InSAR: Tibet and the Three Gorges Dam; extra "(" to be removed
[50] title incomplete; journal, other bibliographic details missing (doi: 10.3390/rs11222670)
[59] title or journal missing?
[60] do not list full given names (Fred, George, Pornpimol) for authors
[61] journal, bibliographic details missing
Author Response
Please see the attachment.
Author Response File: Author Response.docx
