the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 11 Nov 2024
Origin of Changbaishan volcano inferred from simulation of the Cenozoic Pacific plate subduction using geodynamic models with data assimilation
Abstract. The Changbaishan volcano has been considered a giant active intraplate volcano with hidden potentially disastrous eruptive risks, so its origin has attracted widespread attention from all over the world. However, this issue has not been adequately settled down due to its complexity. So far, three primary conceptual mechanisms have been proposed based on seismic tomography and geochemistry. All three mechanisms have been considered to be correlated with the subduction of the Pacific plate. Therefore, we use the best-fit thermochemical geodynamic model with data assimilation, which was determined by tracking the seismically inferred structure of the subducted Pacific slab beneath the Changbaishan volcanic province (CVP), to assess the their relative significance. The findings suggest that the super-hydrous melts in the mantle atop the Pacific slab, resulting from the slab dehydration in the mantle transition zone (MTZ), may primarily contribute to the volcanism of the Changbaishan volcano. Meanwhile, the other two mechanisms, the upward escape of the entrained oceanic asthenospheric material as well as the piling up and thickening of the subducted Pacific slab, may play secondary roles.
- Preprint
(1295 KB) - Metadata XML
-
Supplement
(2106 KB) - BibTeX
- EndNote
Status: open (until 22 Jan 2025)
-
EC1: 'Comment on egusphere-2024-3219', Juliane Dannberg, 11 Nov 2024
reply
Dear authors,
I had a look through your manuscript before sending it out for review, and I already wanted to bring up a few points that would need improvement before I can consider the manuscript for publication.
Specifically,
(1) There are several statements in the text that contradict the current view in the field, without justification being given. If this study makes assumptions deviating from the classical view, there needs to be strong evidence supporting that.
(2) The current state of knowledge is not represented well by the references given in the manuscript, and key references are missing. I found this to be the case throughout the text, including general concepts such as geodynamic modelling of volcanoes and dehydration of slabs, data assimilation, viscosity laws, and the subduction history of the Pacific plate.
(3) There are several statements in the text that are confusing, or where key information is missing. This makes it very difficult to follow both the arguments for the proposed hypothesis and the setup and execution of the modeling study. This information is needed to judge if the proposed model is appropriate to address the question the manuscript poses.
Examples include:“Therefore, some researchers suggested that the subducting Pacific plate entrains enormously its underlying asthenospheric material into the MTZ”
If the slab is stagnating in the MTZ, the asthenosphere can not be below.“utilizing a four-dimensional mantle convection model”
What is a four-dimensional mantle convection model?
(In addition, this paragraph also does not accurately represent the literature; many geodynamic models do suggest a mantle plume as the cause for Yellowstone volcanism.)“The amount of water dehydrated from the subducting plates at deep mantle is the primary factor influencing the magma volume and eruptive intensity of the Changbaishan volcano.”
There is no reference for this. How do we know this?“The thermochemical mantle convection is governed by the equations for the conservation of mass, momentum, and energy”
Is the model compressible or incompressible? Does it include adiabatic heating/shear heating? This information is key, for example to understand if the CMB temperature of 2500 degrees C includes an adiabatic gradient or not (if the former, it seems too low).The description of the oceanic plate composition is confusing: It sounds like it contains oceanic crust and a buoyant layer, but the buoyant layer is also described as crust.
Equation (1): The gas constant seems to be missing here. Furthermore, activation energies of 17 to 42 kJ/mol seem more than an order of magnitude too low for the mantle.
“Considering that the Pacific plate started to subduct between 25 and 20 Ma”
I think the whole tectonic setting requires additional justification. All the plate reconstructions I have seen show the Pacific Plate starting to subduct much earlier (with the exact time depending on the part of the plate boundary and the plate reconstruction). In addition, I would have thought the subduction of the Philippine Plate would play an important role as well.I am confused by Figure 2:
Why is the whole CMB covered by material that is ~1000 K colder than the surrounding mantle? Why is there no hot material? In addition, the colorscale is not perceptually uniform and therefore makes it hard to judge temperature anomalies. Furthermore, the slab deforms a lot, so I would recommend to show the viscosity so that the reader can understand why this happens.What water content in the mantle is implied by a 25% wt fraction in the melt? Is the amount of water consistent with the amount of water that could dehydrate from the slab in the MTZ?
There is no discussion of model uncertainties or limitations.
Best regards,
Juliane DannbergCitation: https://doi.org/10.5194/egusphere-2024-3219-EC1 -
RC1: 'Comment on egusphere-2024-3219', Anonymous Referee #1, 18 Dec 2024
reply
The manuscript by Zhu et al. studies the origin of Changbai volcano based on geodynamic models that are constrained by plate motions, investigating the role of various mechanisms that have been proposed in the literature. The research question is timely and relevant. However, the methods and results of the study are extremely difficult to follow. Therefore, I am unable to assess the conclusions of the study, and thus recommend rejection of the MS in the current form.
The methods heavily rely on the reference “Zhu (2024)”, which is a manuscript currently in review, but I have not been able to find a preprint online. The authors of the current MS only explain the methods of the current study briefly; the reader is referred to Zhu(2024) for details. Also, the current MS relies on the “best-fitting model” from Zhu(2024) and does not explore any controlling parameters systematically. All this may be fine if Zhu(2024) were available, but it is not (as far as I can see). Therefore, it is impossible for me to judge the robustness of the methods, approach, results and conclusions.
For example, the plate reconstructions that are used as a top boundary condition need to be explained, and figures added that visualize the top boundary condition (in map view) for different snapshots should be added.
One of the few things that the authors try to explain in the methods (i.e., in lines 118-124 and onward) is the compositional make-up of the plate/slab. However, this remains also highly unclear. It is unclear what “the former” and “the latter” actually refer to. How thick are these layers and which one is located at the top and which one below? The former layer is described to mostly assume high viscosities and a neutral buoyancy, but nevertheless meant to act as a sticky air layer. A sticky air layer usually has a very low viscosity everywhere and a density of zero. So, this treatment does not seem appropriate. Also, a sticky air layer needs to be sufficiently thick to be resolved by the grid (unclear if indeed the case). A sketch of the model setup would help. Maybe it is included in Zhu(2024), but I do not have access to it.
The values for activation energy look very low. This may be because the ideal gas constant is not included in equation (1). It would be better to include R in the equation. As is, the units don’t work out in the equation. If this is a typo, and the authors just forgot to include R in eq. (1), there needs to be a justification for using such low activation energies.
The results are unclear as well. The quality of the figures is low, and the text does not seem to be supported by the figures.
- In the figures, deltaT is plotted, but it is not clear how deltaT is exactly defined. The colorscale is chosen such that essentially only one isotherm is visible. This needs to be improved.
- In the current figures, the slab does not look like a slab. Thus it is difficult to get an intuition of the dynamics. In Fig. 2a-2b, the slab is disintegrated in many small blobs. Is this a plotting error or a numerical artifact, or even small-scale convection? The location of hot T anomalies is actually consistent with those expected for thermal overshoots (i.e., a numerical artifact), so indeed there may be some numerical problems. Otherwise, where do positive thermal anomalies actually come from? What is their physical origin? The authors should investigate this. If there are significant thermal overshoots in a model that tries to quantify melting, these results would be unpublishable. But possibly I’m getting things wrong and there are no overshoots (Hard to say, as I do not understand what is exactly plotted, and do not know the initial condition).
- Later on (Figs 2e-f, Fig 3 bottom), the slab looks very thick (i.e., very untypical for self-consistent subduction models), perhaps because it is actively “pushed” into the mantle? Or again a plotting issue?
- Figure 3 caption is totally unclear to me. It says melting would be plotted in orange, but the colorscale is the same as in Figure 2. Actually, I suspect that the figure is completely misplaced; it looks like a zoom-in into Figure 2 even though there are some minor differences in terms of details.
- Importantly, the description does not match the results as represented in the figures. For example in line 200~, the authors mention upward flow through slab gaps, but I cannot see any of this in Figure 2. Upward flow through slab gaps is not obvious in any of the panels, so it remains unclear how the authors reach this conclusion. This is just one example. Overall, the interpretations are not well supported by the results, because the results are not clearly presented
- Figure 4 caption is also not very informative (which case? Which W_H2O?). Without detailed description of the plate reconstruction model used and details of the model setup, the robustness of the spatial relationship to the CVP is very difficult to assess.
The figures plot only one cross section through a complex 3D geodynamic model. So even if the quality of figures were improved, more figures need to be added to give the reader an understanding what’s really going on in the 3D models.
I think there is also an issue with the parameterization of hydration, which affects melting in the models and the conclusion of the study.
- Olivine can take up to ~0.5 wt% H2O, depending on p and T. Melt can take significantly more water. The values for W_H2O in the manuscript are on the order of 10s of %, so I think the authors are referring to water in the melt. This is consistent with their use of equation (12), which is based on an equation for saturation of water in a melt. However, in line 216-217, it seems as if the authors are referring to water contents between 0% and 80% in the mantle. Such high water contents in the solid mantle are unrealistic, even if hydrous phases, which are not necessarily stable in the mantle wedge, are considered. In fact, it seems as if the authors do not really distinguish between water in the mantle and water in the melt.
- This is a problem, at least if the authors want to predict volumes of melting. The concentration of water in the melt depends on that in the solid mantle (e.g., olivine) and the degree of melting. It will only be very high at very low degrees of melting. Therefore, fixing W_H2O for a given model does not make sense. The authors would have to make assumptions in terms of the water content of the mantle (also potentially assume different water contents in different regions, such as higher water contents in the slab than the asthenosphere), track water content, calculate melting and track melt depletion, etc. There are many modelling studies that did that properly, and explain how this can be done.
- If the authors instead just aim to predict which areas of the mantle may be above the solidus (basically as a post-processing step), they can do a similar approach as they did. Nevertheless, they need to be clear what the water contents they are reporting actually refer to. Certainly they do not refer to water contents in the mantle. At a given degree of melting, the water contents in the melt can be related to water contents in the mantle source.
If the authors need such high water contents in the melt (do these agree with petrological observations in CVP???), it seems that their preferred model cannot actually well explain the occurrence of volcanism in the CVP. Their model may be missing an important process, or the lithosphere is too tick (I suggest to compare with geophysical estimates), or something else. Thus, I’m not sure if the results support the conclusions.
A discussion is needed to evaluate the limitations of the general approach of pushing slabs into a mantle instead of modelling self-consistent subduction. For example, it is possible that slabs would like to sink fast into the mantle (just based on a low viscosity of the mantle and a high negative buoyancy of the slab), but is not allowed to because it is constrained by a slow plate motion. In turn, if slabs are actively pushed into a high-viscosity mantle, they might undergo natural thickening or underplating. Is there a body of literature that has investigated the limitations in this regard?
Citation: https://doi.org/10.5194/egusphere-2024-3219-RC1
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 87 | 22 | 8 | 117 | 12 | 5 | 4 |
- HTML: 87
- PDF: 22
- XML: 8
- Total: 117
- Supplement: 12
- BibTeX: 5
- EndNote: 4
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1