Electrical conductivity in the mantle transition zone beneath Mongol-Okhotsk suture: revealed by the geomagnetic signals of ground observatories
Abstract. The closure of the Mongo-Okhotsk ocean has a strong influence on the tectonic evolution of Northeast China. However, the dynamic mechanism in the Mongol-Okhotsk suture area is controversial. This paper intends to obtain the deep structure of beneath Northeast China based on geomagnetic depth sounding, and constrain the subduction of Mongol-Okhotsk Ocean from the perspective of electrical properties. This paper collects and processes the data of geomagnetic stations in China and adjacent areas, and obtains stable C-response data. The staggered grid finite difference method is used for forward modeling, and the finite memory quasi Newton method based on L1-norm is used for inversion. The three-dimensional inversion of geomagnetic data is carried out in spherical coordinates. The intensive model testing stations can obtain high-resolution underground electrical structures. The measured data show that there are obvious high conductivity anomalies in the mantle transition zone in Northeast China, especially in the west of the Great Xing’an Range, showing an area of high conductivity anomalies. Combined with the regional tectonic background of the region, we speculate that the high conductivity anomaly body is related to the southward subduction of the Mongol-Okhotsk Ocean. The Mongol-Okhotsk Ocean subducted under the Eurasian plate at a small angle in the southward direction. With the closing of the Okhotsk Ocean and the extension environment after the termination of subduction, the subducted oceanic crust plate has been faulted and depressed and partially stopped in the mantle transition zone.
Yanhui Zhang et al.
Status: open (until 07 Jul 2023)
- RC1: 'Comment on egusphere-2023-480', Anonymous Referee #1, 03 May 2023 reply
Yanhui Zhang et al.
Yanhui Zhang et al.
Viewed (geographical distribution)
This manuscript aims at constraining the 3-D conductivity in the MTZ by using long-period geomagnetic variations measured at a number of ground observatories.
In my view, the manuscript cannot be considered for publication in its present form. There is one fundamental issue, which I will discuss below, but first some less critical, but imnportant things:
-- No information about the data is given. Authors claim to have used observations from 30 locations, but none of the figures shows where they are located. Let alone things about how data is measured, whether it is open and available...
-- No information about data processing were given. Which core-field model was subtracted? How they deal with long-term biases and trends. Were time gaps filled or not... Responses are shown only for a few sites. No R^2 coefficients or any other statistics on the quality of the responses are given.
-- Only final model is presetned. Its stability and the choice of many underlying hyper-parameters are fully ignored. What starting models, regularization parameters, etc were tested? How robust are model features subject to these choices?
All points above could have been fixed within a major revision, but the critical weakness of the study that, in my opinion, makes it unsuitable for publication is this:
Authors used C1 responses, derived as a ratio of Br/Btheta (also known as Z/H method after Banks 1969). The key limitation of this type of responses is that it is valid ONLY if your source is described by the single P10 harmonic. We know from data it is never really true. Yes, it is a dominant sources during the main phase of magnetospheric storms, but these phases are short. Assuming 100% P10 source for years (sometimes decades) of data time series is really not justified and should not be done. The problem is that once you make this assumption (and you always make is if you use Z/H method), your responses will be affected and biased by non-P10 effects. These source effects will then propagate into your subsurface conductivity model. The non-P10 source effects are way stronger than plausible EM induction effects due to 3-D mantle anomalies. Given that your 3-D problem is terribly underdetermined and non-unique, there is no way to discriminate between source effects and real anomalies.
There is >20 years of research in the field of global EM induction that went into trying to overcome limitations of this old C1 responses, and derive more elaborate transfer functions that can handle complex sources. Authors completely ignored this body of knowledge and new methods.