Electrical Conductivity Evidence for the Existence of a Mantle Plume Beneath Tarim Basin

Round 1

Reviewer 1 Report

Review comments on

“Electricity Evidence for the Existence of Tarim Mantle Plume”

General Comments

Basically a well written paper on an interesting topic. The paper investigates using Geomagnetic Depth sounding to explore a region in China and has identified an anomaly in this area and is validating with a new 1D SA inversion.

One major issue I found was that the soundings look very deep into the earth for a local anomaly using a 1D analysis. At this depth and this size why does a 1D code provide a useful result? I would think that a 2D or 3D code would be much better. At the very least perhaps run a few forward models to see what a 1D inversion would return.

Other than this question and a few minor adjustments it is ready to go

Details

In Figure 2 I would like to see the time series plots with the long period drift removed so that we can data coherency between the stations.

The inversion uses a model where the layers are fixed and only invert for the conductivity. How is this justified? How were the layer depths and thicknesses arrived at ?

P15 The insensitivity to Gaussian noise is one thing but a lot of background noise is not random. Data bias is much tougher and also tough to estimate.

Final comments. A pretty well written paper and a fair useful result. Accepted with only minor adjustments.

Author Response

Dear editor and reviewer:
Thank you for your letter and reviewers' comments on our manuscript entitled "electrical evidence for the existence of the Tarim mantle plume". These comments have important guiding significance for our research.
1. The data measured by geomagnetic stations are mainly affected by the anomalies below them. When the horizontal distance between the local magnetic station and the anomaly is far, the geomagnetic station will not observe useful information.
2. 3D or 2D inversion is indeed a better choice, which will be our next research point.
3. We think that the coherence between stations can be directly reflected by C-response. The time series without long-term field will not provide better information than C-response.
4. The model of fixed layer thickness is based on the division of important underground interfaces. With the deepening of the depth, the phase transition of matter will change its physical properties (conductivity). In this way, the average conductivity of different rocks in a certain depth can be well calculated.

Reviewer 2 Report

As I already said, I cannot assess the method, just the general context of mantle plume.

I think "thousands of kilometers" diameter (as stated on p. 1) for a plume is unrealistically large, especially in the transition zone, so I don't think a 1-D model is appropriate. It is not shown that it might still be appropriate.
Besides, more generally, I wonder whether a 1-D model for the transition zone (410-660 km deep) is appropriate when some of the stations are much less apart. For example, HOQ and WMQ are only 200 km apart. So surely, lateral variations must play a role and I am not convinced that the 1-D approximation is appropriate. At least, I didn't see it justified anywhere.

Besides, if there is a Tarim LIP from the early Permian, you don't necessarily expect the plume is still underneath because the plate may have moved. It is not shown that within the uncertainty of plate reconstructions, it is realistic that the plume is still underneath. And if it was, why would there be Permian Volcanism and nothing afterwards?

Besides, this only uses EM. But conductivity may be due to both temperature and water content. So it would be good to use two methods (such as conductivity and seismology) to constrain two parameters (temperature and water content). But this is not done. Also, there isn't any discussion of whether there is any other evidence (from other fields) for or against a Tarim mantle plume presently existing, which is a glaring omission. For example, at the very end of the discussion they mention how much transition zone thinning is expected, but there is nothing about whether any of this is actually observed.

Page 2: "conductive convective fluid in the core and the mantle". Surely, in the mantle it is also the solid part that is conductive. Especially, if you want to infer temperature, the temperature dependence of the conductivity of the solid should also matter.

Fig. 1, horizontal axis should be "°E".

p. 11 "Within 410 km" should better be "Above 410 km"

In Figure 7 why would WUS and WMQ which are furthest apart both be affected by the plume whereas the two inbetween wouldn't?

Besides, if there is an effect of a plume, we would expect a localized high conductivity at all depths, not a widespread (horizontal, 1-D) anomaly at a specific depth range such as 660-900 km. For example, at depth below 900 km, the conductivity is much lower than global average, but this result is then just brushed away by saying the recording time was not long enough. But the recording time was about 10 years (3*10**8 s), so from all I know (as said, I am not an expert) this should be long enough. For example, Deschamps and Khan (EPSL 450, 108-119, 2016) write that "Measurements at longer periods, up to about 2 years, would be needed to sample the lowermost mantle (≥2500 km)". So I think this excuse is not valid. Or it is just stated that "conductivity will decrease in the lower mantle", without further explanation -- see below.

p. 16, 150 K, 450 K, 650 K add "anomaly" to distinguish from absolute temperature as in Fig. 13.

Fig. 13 caption, should be "conductivity at 660-900 km"

The discussion in the context of mantle plume is all muddled up. Firstly, the temperature at the core-mantle boundary is not 2550 K, but much higher. 2550 K is approximately the temperature you get from extrapolating the adiabat to the CMB, but the temperature jump across the thermal boundary layer at the base of the mantle has to be added to that. Then they seemingly claim that the difference between 2550 K and 1925 K average temperature at 660 km would cause a ~650 K anomaly, but obviously, if material rises adiabatically, it would also cool accordingly. So, in order for material to have an anomalous temperature at 660 km, it would already have to be hot, compared to the adiabat, when it starts off near the core-mantle boundary. What is meant by the orange dotted line "on the far right"? I don't understand how the estimate of r/a between 0.4 and 0.6 was obtained. They say it is according th the calculations of Shunichiro Karato but Reference 52 is supposedly a book by Todo. Besides, I would prefer some original reference which I could look up, not a book where I don't have access to. Besides, I don't understand why this information is relevant here (it is not explained). And why would temperatures at the bottom of the lower mantle decrease at the edge of a plume? It is not explained and doesn't make any sense. Besides, the results shown only go down to ~1150 km, there is nothing shown for the bottom of the lower mantle.
Basically, it gives the impression that the authors don't even understand the basics of mantle plumes, and mantle dynamics in general, and are also not capable of putting their results in a wider context.

Author Response

Dear reviewers

Thank you very much for your valuable advice. We have read your suggestion carefully and revised the article.

1. We corrected the mistakes in the article (Fig. 1, P. 11, P. 16, FIG. 13)

2.Because we are doing one-dimensional inversion, it is difficult to determine the exact location of the anomaly. Therefore, WUS and wmq may be affected by anomalies. We will collect more station data. 3D inversion will take some time to complete the next step.

3. In the new draft, we give the evidence of seismic imaging to prove our point.
4. There are many factors that affect the detection depth of GDS method, and the continuity of time is the key factor. The duration and continuity of the data measured by the four stations are not enough to reflect the deep and reliable conductivity.
5. We rearranged the conclusion to make it more organized.

Kind regards,

Author Response File: Author Response.doc

Reviewer 3 Report

The authors propose to use simulated annealing (SA) calculation to perform one- dimensional inversion of Geomagnetic Depth Sounding (GDS) to obtain the conductivity information of the lower mantle beneath the Tarim area, and calculate the temperature of the lower mantle according to the relevant formula of the petrophysical experiment, and give evidence of the existence of the Tarim mantle plume. I read the paper with interest, it reads very well, the results look encouraging and well presented, the English is fine with me. The manuscript organization looks very good and all presented materials are suitable and correctly discussed and presented. Therefore, I recommend to accept the paper without any further review.

Author Response

Dear editor and reviewer:
Thank you for your affirmation of our manuscript entitled "Electricity Evidence for the Existence of Tarim Mantle Plume". This will encourage me to conduct a more in-depth study of the Tarim Region.

Reviewer 4 Report

In this manuscript, simulated annealing algorithm combined with statistical analysis technology is proposed to perform one-dimensional inversion of four stations in Tarim area. It would be interesting to the readers. However, the experimental results should be improved. I suggest that the section 3. "Results" and section 4. Discussions can be merged and reorganized. One is clean data and another is noisy data. For noisy data, the author assume that the obtained noise type is Gaussian noises. Why? White noises? or color noises. Why 5%, 10%, 12%? Generally, the added noises can be measured as dB not 5%, 10% and 12%.

Author Response

Dear Editors and Reviewers:

Thank you for your letter and reviewers' comments on our manuscript entitled "thesis title". These comments have important guiding significance for our research.
1. After the test in Section 3, it has been verified that the classical simulated annealing algorithm can retrieve the clean data, but the classical simulated annealing algorithm can not retrieve the data with noise, so an improved scheme is proposed.
2. Most of the noises that affect the data are Gaussian noise, such as cosmic noise. After a lot of tests, excessive noise will make the inversion results no longer reliable, so the author limits the noise amplitude to 12%.

Round 2

Reviewer 2 Report

I think there is still considerable room for improvement. They haven't even attempted to discuss some of the things I requested.
--> Do we expect the plume that created the Tarim LIP to be in that area? That point would need to be addressed with plate reconstructions. Otherwise, there might still be a plume but it might be unrelated to the Tarim LIP, and proximity might be fortuitous.
--> At what depths and for what lateral extent of the anomaly is a 1-D method appropriate?
I still find the geodynamic explanation unconvincing. I think one scenario that would sort of make sense is
--> there is a plume rising from the lower mantle, but because it is too narrow to be detected by this 1-D method, it is not imaged at depths greater than 900 km
--> The endothermic phase transition at depth 660 km hinders the plume from penetrating from the lower to the upper mantle. Therefore, plume material spreads widely between the phase transition, and can therefore be imaged with the 1-D method in the entire area.
--> From this spread-out anomaly, the plume penetrates beneath WMQ into the upper mantle. This depth is sufficiently shallow (and the anomaly is sufficiently extended) that it can be imaged with the 1-D method.
So it could be a plume, I suppose, but to call this "evidence" I still find a fairly strong statement in the title. At least I am not convinced.
Unfortunately, the seismic model shown doesn't really support any of this either. It is at 39° latitude, which is further south than all your stations, and doens't show any extensive slow anomaly in the uppermost lower mantle, which would be supporting your model.
Also, I still think the ratio of plume tail diameter to plume head diameter has nothing to do with this. I still cannot access that book, but I think it refers to the head and tail diameter while the plume is rising through the mantle. In contrast, it seems like they interpret it here as the ratio of the plume tail in the lower part of the mantle to the upper part of the mantle at present, i.e. long after the plume has reached the surface ~290 Myr ago.

Author Response

Dear reviewer:

Thank you very much for making us aware of the shortcomings of our article. This makes us realize that the Tarim large igneous province has little connection with the high temperature anomaly we analyzed. The formation position of Tarim large igneous province is not the present position.
1. After the major revision of the paper, the thermal anomaly will not be related to the large igneous province.
2. Another section in Huang's article is quoted by us, and we think it can have a good correspondence with our results.
3. According to your suggestion, we analyzed the conductivity and temperature of the mantle transition zone. In addition, the significant increase of water content in the mantle transition zone is usually related to the dehydration of the stagnant plate, while the stagnant plate has not been clearly found in the Tarim transition zone, so it is necessary to explain the high conductivity anomaly in this area through the increase of temperature.
4. We modify the title to make the expression more accurate.
5. You give us a perfect suggestion to revise this article. And your hypothesis is very important to us. At the same time, we also realize that the ratio of the head to the tail of the plume is meaningless.

If there are any deficiencies, please give your valuable suggestions. We would appreciate it.

Kind regards,

Junhao Guo

Author Response File: Author Response.doc