Talk:Seismo-electromagnetics

Reference format

It would be a lot of work to do a great job on this article. But just to get started, what I've done is to cobble together some material from old versions of the EQ prediction article, together with materials taken from the ledes of VAN method and Quakefinder.

Some of this material is using the 'Harvnb' reference template, which gives short-form notes and requires separate complete citations. Articles using this reference format typically put the complete citations in a section called 'references' which is separate from the 'notes' section.

Other material uses the 'cite' template variants, which place the entire citation in the 'references' section. This 'references' section is essentially equivalent to the 'notes' section in a 'Harvnb' formatted article. This article was initially in this latter style.

In copying material from other articles using the 'Harvnb' template, the complete reference materials didn't automatically get copied. Bringing those references over is a painstaking process, which I haven't yet begun.

My experience across Wikipedia is that most articles use the 'cite' templates, which are much easier to work with. My plan is to keep this article in this format as I bring the full citations in. Comments or objections? JerryRussell (talk) 22:11, 29 December 2016 (UTC)

You're confused. But understandably so, as usage, and verbiage about usage, is confused all across WP. Allow to me offer some points of clarification.

• The {{harv}} family of templates produce short cites (such as "Smith 2004") which will auto-magically link to a full citation (if the latter is properly prepared).

• The {{cite xxx}} family of citation templates, and the comparable {{citation}} template, produce full citations, "full" meaning with complete bibliographic details, in some kind of standard format. They vary in the details, the most important for us being that {cite} requires an additional |ref=harv parameter to use it with {harv}.

• The prevalence of {cite} templates has nothing do with ease of use, and everything to do with people using what they have already seen used, and in the same manner.

• Neither {cite} nor {citation} "place the entire citation in the 'references' section." Citations, whether full or short, are placed by the editor. In regard of placement an editor has a series of choices:

• First, and foremost: whether the citation to be placed "in-line" in the text should be a) a full citation (using {cite} or {citation}), or b) a short cite (using {harv} templates). This has major consequences, which I will touch on below.

• Second: whether the "in-line" citation be displayed in a) the displayed text, or b) in a note that is displayed elsewhere. If the latter, then the in-line citation is placed between <ref>...</ref> tags. Note that when the first decision is to use a full citation in-line, it is invariably placed in a note, as displaying all of those bibliographical details in the text tends to confuse and annoy the readers.

• Third: where to place the {{reflist}} template, which is where the notes (that is, the material from the <ref> tags) is displayed. This is usually in a section named "Notes" or "References". But note: this placement, and the name of the section, have nothing to do with whether short cites were used, or not.

• Fourth (but only if full citations were not placed in-line): where to place the full citations. Usually they are collected together in a section titled "References", "Sources", or something similar, but there are other possibilities. E.g., it is possible to collect the full citations of the sources used in each section at the end of each section.

You said that use of {Harv} "requires separate complete [i.e., full] citations", but this is a bit misleading. More accurately, use of {harv} enables separation of the full citation from the specific use in the text (and possibly at multiple locations). More importantly, it enables linkage to a given full citation from multiple locations, without having to replicate the citation at every location.

Having the full citation (complete with ALL the bibliographical details) embedded in the text can be handy when copying text to other articles. But it can also be a pain if you delete a citation that is referenced from elsewhere. Having citations mixed in with other notes and explanatory material can be confusing (which is why some articles have parallel sets of notes), and you can't sort them (the notes are presented in the order the software finds them) or give them special treatment. It is also very annoying trying to edit text with all those citation templates present, and also annoying trying to check and clean-up citation templates when they are scattered all throughout the text. If you work on only short articles, where sources are usually cited only once, and are not so keen on providing in-source page numbers, then these problems are not so big. But in even moderate-sized articles it is generally more advantageous to use short cites for in-line citation, and have the advantage of separating bibliographic detail from the text.

~ J. Johnson (JJ) (talk) 23:54, 30 December 2016 (UTC)

JJ, thanks for the clarifying information and terminology. I had only been seeing the downside of the short cite format and {harv} templates, but now I understand that they also have advantages. JerryRussell (talk) 00:10, 1 January 2017 (UTC)

Created the 'sources' section, and copied in the sources from the 'VAN method' and 'EQ prediction' articles. A remaining task is to convert the rest of the references in the article to {harv} format. JerryRussell (talk) 01:09, 9 March 2017 (UTC)

Lede revision

Generally I think it looks good, and gives a better view of the reasons mainstream seismologists are skeptical about seismo-electromagnetics as a means of EQ prediction. One question I have is about the phrase in the opening sentence, possibly associated with seismic activity. Is there any question about this? I thought it was an accepted fact, that there is electrical activity associated with earthquake events and volcanic events. The debate is about whether there is any precursory activity. But if there is sourcing indicating any doubt about the simultaneous electrical activity, I would withdraw this question. JerryRussell (talk) 22:26, 7 January 2017 (UTC)

No need to withdraw the question; I'll just answer it. Which is: definitely! A big problem with all claimed precursors is the claimed connection between some event and some subsequent quake. I don't know of any alleged precursor for which there is a record long enough to show any kind of association. Many claims (such as the Corralitos event) are single instances, and the alleged association is entirely one of time and space coincidence, with no showing of any underlying connection. I find it just amazing that some practicing scientists are so quick to make these associations. You might note that VAN did not have any workable theory to explain their SES, just the result of lab experiments on dry rock, which they assumed applied also to the earth's crust. (Which is rock, right?? So there!) When challenged they suggested differences in geology, movement of fluids in the crust, etc., but all rather ad hoc and never as a formalized theory that could be tested. Their fame (and such potential credibility as they had) was based entirely on the predictions. (With proof of a method in hand, they could sort out the theory later.) Gruznow raised the objection that VAN had not excluded industrial noise. But what really hulled them was when Pham et al. came out with pairs of broad-band EM sensors, and triangulated several SES-like signals (what a bold idea!) to .. irrigation pumps, an electrical transmission line, and so forth. So while there are some reports of coseismic EM effects, the events claimed as preseismic seem to be entirely coincidental. ~ J. Johnson (JJ) (talk) 00:06, 8 January 2017 (UTC)

JJ, my question is whether there are observations of electromagnetic activity during the earthquake or volcanic event. I understand that the existence of any precursor activity is highly controversial. But, the study of seismo-electromagnetism would also include EM simultaneous with the quake, right? Not that the article currently says a single word about that, in its current state. But I've seen a paper or two that mention the simultaneous activity, and seem to take its reality for granted. JerryRussell (talk) 02:04, 8 January 2017 (UTC)

Here's a book chapter by Johnston that says causal relations between coseismic magnetic field changes and earthquake stress drops are no longer in question. This is a big survey article, with plenty of specific data.[1]

Strangely, Susan Hough's book says: Magnetic instruments can and do record earthquakes. But upon close inspection, one finds, without exception, an absence of true co-seismic electromagnetic signals. (p. 137) Which begs the question: what does she mean by a "true" co-seismic signal, as opposed to one which is merely recorded by a magnetic instrument? JerryRussell (talk) 20:43, 14 January 2017 (UTC)

Regarding "true" co-seismic signals, you have pretty much answered it yourself: "as opposed to one which is merely recorded by a magnetic instrument". Note: the recorded by attests only to a certain functioning of an instrument, which, based on certain engineering concepts, is expected to behave in that way in the presence of certain physical phenomena. As you read in §2 of Johnston (2002), "data up through the mid-20th century" are suspect because the instruments were susceptible to the physical displacements due to earthquakes. (Which is to say: whoops! there is another way of getting that result.) Note also his figure 1, showing the sudden drop in the magnitude of reported tectonomagnetic anomalies after the development of the proton precession magnetometer (which is not sensitive to physical displacement). It appears the device used by Fraser-Smith was susceptible to such displacements.

But even where instruments function as expected, and the results taken as corresponding with actual physical phenomena, there is still another question: is that phenomena actually seismic? Remember that instances of VAN SES were traced to irrigation pumps, etc. (Hough alludes to this on page 139.)

You seem pretty impressed with Johnston's review, but note that it was in 2002. More pertinent are his 2006 paper and Park's 2007 paper, both on the Parkfield quake, where they did record co-seismic signals in a manner sufficiently rigorous to make Varotsos et al. look pretty bone-headed. But note: absolutely no trace of pre-seismic signals. And if a phenomena is not pre-seismic it is hardly a useful precursor.

Our interest in co-seismic activity is mainly in understanding the means the SEM crowd claims generates pre-seismic signals. Their claims are also dependent on some usually unstated postulates, which Hough mentions on page 137. ~ J. Johnson (JJ) (talk) 22:36, 15 January 2017 (UTC)

EM precursors + induced seismicity

1. SES (up to 3 months before the eq) - similarity to heart attack
2. MHz (up to several days before the eq)
3. MHz+kHz (up to several hours before the eq)
4. kHz (several hours to a few hours before the eq) - similarity to epileptic seizure
5. EM silence (a few hours before and up to the eq) - no turning back, earthquake follows
6. Earthquake

EM modeling

• Electric, elegtromagnetic and magnetic signals may be detected. Electrics reach the surface as shown in the SES, electromagnetic RFs can surface for M>6 earthquakes (not if the focal area is below the sea), magnetics can travel better through water.
• Ionospheric disturbances may occur before the eq. It has been observed that the height of the ionosphere can both lower or raise before an eq. Micro-crack distrubution producing micro-em-pulses over the large area (hunderds of km) stimulated, behaves like a fractal antenna (extremely powerful).
• Infrared radiation observed by satellites (raise of temperature) may occur before eqs, surrounding the focal area.
• Induced seismicity is also part of this article. Α , Β

-AA-85.74.33.233 (talk) 22:28, 5 March 2017 (UTC)

Hello AA, I've been trying to read your first reference here: Donner, Potirakis, Balasis, Eftaxias & Kurths 2015. I see that they describe the basic sequence of MHz, KHz and silence that you mention here. The description is based on a literature review. They say that the MHz is due to cracking around the stressed fault, KHz is due to slow stick-slip like stage, with a tri-critical crossover leading into a first-order phase transition, and that the quiescent phase is due to a dynamical slip and super-shear mode. (I am parroting their words, more or less, with no clue what they mean. But the analogies to stroke & heart attack make even less sense to me.) The remainder of the paper is focused on an examination of the third-stage KHz EM signals from three earthquakes. They find that eigenvalue entropy increases and then suddenly drops as the EQ approaches. Then at the end of the paper, there is an important confession: Due to the limited amount of data considered here for the pre-mainshock period, we cannot yet draw any final conclusions about whether or not the herein revealed complexity and eigenvalue entropy signatures are truly unique to the preparatory stages of major EQs. They also mention that one of the EQ they studied was significantly different from the other two.

In other words, they haven't analyzed a lot of data from their sensors that isn't associated with any EQ, so they have no idea how unusual these KHz signals and entropy signatures are?? Am I reading this correctly? As JJ would say, where's the beef? JerryRussell (talk) 05:42, 7 March 2017 (UTC)

There is no beef here, only cabbage and broccoli. SV1XV (talk) 06:12, 8 March 2017 (UTC)

Also, AA, in order to include any of this in the article, we would need to establish basic noteworthiness under WP:DUE. There needs to be some independent secondary source review of the basic theory that EQ are marked by MHz signals, then KHz, then pre-quake radio silence. I know that Eftaxias has several co-authors, but their independence tends to be at least open to challenge. It can be hard to trace where the money is going behind the scenes.

In academic circles, I believe it can also be pretty easy to get professors and students to sign up as co-authors to contribute their expertise regarding some obscure mathematical or technical aspect of such a complex paper. They don't necessarily review, or stand behind, the idea that these signals are real. If any of the co-authors are uncomfortable with this, they can point to the disclaimer in the paper.

The best reference to give, would be some review article or endorsement from a mainstream seismologist. You came up with such a source for Natural Time, which was very helpful. Even a disparaging reference, or a discussion in a reputable popular magazine or journal, could help. Do you know of anything? JerryRussell (talk) 03:47, 8 March 2017 (UTC)

You have been reading the publication the wrong way because I have introduced its model in the bullets nmbered list above. We cannot use the model, as it is only a proposal (although I have yet to dig on endorsments), but we can use what is summarized in the article. What has been shown is well summarized: micro-crack opening produces EM emissions, shown both in the field and the laboratory, in a wide range of spectrum freequencies, and MHz emissions come systematically before kHz. It has also been shown that there is systematical EM quiescence very near the earthquake (this is the main reason seismologists say earthquakes cannot be related to EM). These are the facts of observations, as well as the ionospheric disturbances and the raise of heat detected before an earthquake. We do not study earthquake prediction here, but if and how earthquakes relate to EM. The similarity to biological and other systems show EM emissions observed before earhquakes follow known critical phenomena laws.[2][3][4][5]-AA-85.74.33.233 (talk) 07:16, 8 March 2017 (UTC)

AA: I have introduced its model in the bullets above. You mean your numbered list, rather than the bulleted list? The numbered list seems to be a summary of the intro to Donner et al, while the bulleted list is more of a possible outline for our article here. These are the facts of observations -- yes, but the reader of Donner et al. is presented with something of a chore, to track down all of the cited papers and find those that contain evidence of these observational claims. Then, for Wikipedia purposes, we would need to find independent secondary reviews.

Thanks for the medical papers. JerryRussell (talk) 23:57, 8 March 2017 (UTC)

I have corrected the type of list. What is evident (not a proposal) can be used, we are not doing original research if something has been shown to be as presented. But of course we will have to follow publications cited in Donner et al to document our article. Yes, physicists can save lives also through reasearch for earthquakes, applied in biology. Who doctor knew that a perfect heartbeat means that the patient is about to die... But please stay focused in EM here. Natural EM signals look like biological ones, and this (between others) means they are not man-made.-AA-85.74.33.233 (talk) 00:50, 9 March 2017 (UTC)

Au contraire. The "SES" of VAN look like industrial noise. This was convincingly shown, and VAN never showed otherwise. For all that people keep popping with claims that they saw some kind of EM "signal" prior to an earthquake, no one has shown ability to meaningfully predict earthquakes. ~ J. Johnson (JJ) (talk) 04:35, 12 March 2017 (UTC)

Eftaxias ULF observations

AA inserted some text into the lede about these two papers by Hayakawa, Schekotov, Eftaxias, et cetera. First of all, I believe it's undue weight to mention these primary source articles in the lede. So I've moved them to their own section. But I'm not sure they belong in the article at all, unless there's some independent secondary source review. This is likely to get deleted unless there's evidence of noteworthiness. Also, is it correct to identify Eftaxias as the principal investigator in this project? JerryRussell (talk) 15:43, 9 March 2017 (UTC)

The sources were put in the lede section only to balance the opposite view for no findings in a big earthquake and should not be left there (there is also this publication for a non-notable earthquake and the modeling paper based on 3 earthquakes, two of which are 1999 Athens earthquake and 2009 L'Aquila earthquake). I agree we need endorsments. Eftaxias is the principal investigator of the MHz/kHz/EM_silence model. I see trusting Eftaxias to analyze Kobe, Tohoku and Sichuan earthquakes as an acceptance of his work and presence in the field but I do not know if this is enough for a separate section on his work. I would prefer to build the article by types of EM signals, primarily, and let research groups appear in their own sections later in the article.-AA-85.74.33.233 (talk) 17:53, 9 March 2017 (UTC)

Hi AA, my !vote is still that the dominant view in the current seismological literature is as JJ says: that there's no reason to expect measurable seismo-electromagnetic EQ precursors, and that the findings of the Parkfield experiment can be taken as confirmation that there are no precursors. So, the lede should spell this out. Whereas the findings by Eftaxias are not endorsed by the scientific community, and so they should be represented less predominantly, if at all. If there's not enough information to justify a separate section, there's certainly not enough to justify presenting this in the lede.

I also feel that if Eftaxias and his group are the only ones claiming that there are MHz and KHz signals, it would be undue to organize the article around that theory. It makes sense to me, to continue to organize according to the findings of each noteworthy research group. The other groups all agree these are ULF signals, and they're looking at various detection methodologies and mathematical signal processing techniques, is that right? JerryRussell (talk) 22:21, 9 March 2017 (UTC)

Tagging for noteworthiness. I found a couple of independent-looking references, so now I feel a little better. One of them is an Italian group making similar observations prior to an EQ in Crete, and the other talks about acoustic as well as EM signatures of fracture processes, and uses borehole probes as deep as 2km to look for ULF and KHz signals. Both mention Eftaxias' work, as supportive of their own. Still not sure if this is good enough, so I'm leaving the tag. JerryRussell (talk) 03:27, 12 March 2017 (UTC)

VHF and HF are MHz, VLF and LF are kHz. ULF, specifically in seismology, range from Hz to MHz, by many alternate definitions (see the WP article). ULF is rather confusing to be used as a term. When we read ULF in publications, any MHz, kHz or other band can be behind the term.

Another way to structure sections could be electric, electromagnetic and magnetic variations observed. But this way of sectioning might have more to do with the channel of propagation and the detection organology rather than the mechanism generating the signal, so I recommend we use it for sub-sectioning.

We should be primarily interested in presenting the generation mechanisms proposed (piezoelectric phenomena, micro-crack EM emissions, etc.), or bands or type of emissions, rather than sectioning the article by research group.

In the Parkfield example the analysis done is based on non-modern methods on critical phenomena, so it is expected to miss any signals. I wonder for the findings if Natural Time is to be used on the data available.

There is a difference between earthquake prediction and seismo-electromagnetics. Traditional seismologists do not accept electromagnetics as physical quantities related to earthquakes. We state the traditional seismologists view and then we forget them, except if they answer to publications, with publications. Everything else is proposals on observations, not ways to predict earthquakes, and should be treated accordingly. As nothing is widely accepted, peer-reviewed publicatons should be enough if the publications withstand the test of time, especially without any answer. Verifyability should be enough, 50 or 100 publications is endorsment from at least 100 or 200 review(er)s.-AA-85.74.33.233 (talk) 14:32, 12 March 2017 (UTC)

Bullshit. Though I grant you, it is very fine bullshit, of a higher grade than usually seen. Especially about the "non-modern methods on critical phenomena", "non-modern methods" being a reference to what everyone else does, VAN alone having a "modern method" that unlocks the secrets of all natural phenomena. Not as funny as Varotsos getting his butt served up on a plate when he criticized the Parkfield findings, but still a joke. (I'll dig out a copy if anyone wants to see it.) Your argument that being ignored amounts approval is just absurd. We have good authority (both primary and secondary sources) that VAN has failed, and is debunked; there has been no reason to say more. Despite 30 years of trying AND "modern methods", skillful prediction of earthquakes is yet to be shown. ~ J. Johnson (JJ) (talk) 20:40, 12 March 2017 (UTC)

AA, the Wikipedia ULF article is Ultra_low_frequency. A range of 1 milli-hertz to 100 hertz would cover all the various definitions given. It is definitely not kilohertz or megahertz. I don't find it all that confusing. Your suggestion "peer-reviewed publications should be enough" is not Wikipedia policy; generally, primary research findings are not considered encyclopedic unless they've also been discussed in secondary reviews of some sort. See WP:WPNOTRS. JerryRussell (talk) 21:40, 12 March 2017 (UTC)

"In magnetosphere science and seismology, alternative definitions are usually given, including ranges from 1 mHz to 100 Hz,[2] 1 mHz to 1 Hz,[3] 10 mHz to 10 Hz.[4]" is taken from the article itself. You have just proven confusion, Jerry. Next, if you read carefully, secondary sources are preferred, but primary sources are not forbidden, as long as original research is avoided. OR in a nutshell is: "Wikipedia does not publish original thought: all material in Wikipedia must be attributable to a reliable, published source. Articles may not contain any new analysis or synthesis of published material that serves to reach or imply a conclusion not clearly stated by the sources themselves." When a scientist has 100 publications, accepted by peer review, one after the other, even in Physical Review Letters, what is reliably published out there is well grounded to fit here as a proposal.-AA-85.74.33.233 (talk) 00:40, 13 March 2017 (UTC)

Hilarious. Are all those papers in the past decade? Because anything older than that (as you have stated before) would be "old science", right? And just being published doesn't mean that your claims are accepted. Peer-reviewed means good enough to be presented. In VAN's case there is also a plenitude of peer-reviewed articles rejecting VAN's published claims. ~ J. Johnson (JJ) (talk) 23:41, 13 March 2017 (UTC)

Hi AA, I don't understand what you find confusing about the quoted passage from the ULF article. Do you agree that the mHz abbreviation refers to milli-hertz (.001 Hz) and not mega-hertz (1,000,000 Hz)? Does it bother you that various sources might use the same term differently? My view is that the term is widely used in the literature, and therefore we should use it, even though the exact intended frequency range can vary from author to author.

With regards to your statement above, When we read ULF in publications, any MHz, kHz or other band can be behind the term. I think that's false. It always means a frequency below 100 Hz, which excludes any Kilo-Hertz (KHz) or Mega-Hertz (MHz) band.

I agree that primary sources are not strictly forbidden. I like to include them in articles as part of achieving complete verifiability. But, the guideline says that Wikipedia articles should be based mainly on reliable secondary sources and that Large blocks of material based purely on primary sources should be avoided. I believe there's room for discussion about the line between acceptable and unacceptable. In any particular case, the outcome depends on the consensus of editors. This Eftaxias section strikes me as borderline as it stands. I only spent a few minutes searching for secondary references to the material. I believe there's probably more to be found. I trust you'll let us know if you can find anything else? Thanks. JerryRussell (talk) 01:58, 13 March 2017 (UTC)

As I said above, meeting the minimum requirements for publication does not amount to acceptance. It means being presented for consideration. And many papers are rejected. WP is wary of the use of primary sources because WP editors are almost always unfamiliar with the literature or scientific opinion broadly, and too likely to run with one shiny bit without regard to everything else.

The basis of WP (at least for scientific topics) is supposed to be scientific opinion. Unfortunately, in practice this usually comes down to the opinions of WP editors as to what scientific opinion is. Which in many cases is wildly inaccurate. ~ J. Johnson (JJ) (talk) 00:05, 14 March 2017 (UTC)

The basis of WP (at least for scientific topics) is supposed to be scientific opinion. Is there a policy that says so? WP:NPOV says we are to include all significant viewpoints that have been published by reliable sources, which could include non-scientific or even pseudo-scientific sources, as long as they are reliable. (I confess I still have a little trouble with the idea of pseudo-scientific articles in high-quality peer reviewed journals, but I suppose it's possible.) It's easier for lay editors to identify all the viewpoints, and their prominence in secondary reviews, than it is to sort out what's true and what's false.

Considering the above-expressed confusion between millihertz and megahertz, it's feeling to me like AA might be more of a student or a nationalist fan of Greek EQ scientists, rather than a seasoned expert researcher. If he or she is a student, I hope we can make this a learning experience. JerryRussell (talk) 02:40, 14 March 2017 (UTC)

A "fan" (nationalist or not) is a quite mild way of putting it. He is much deeper into the topic than a student would be, and more "seasoned" than open-minded; his persistence suggests a deep or strong connection, part of which can be traced back to the UOA, so most likely much closer to the principals than a mere fan. (And I'm ignoring that milli/mega confusion. Even for wasting time I've got better things to work on.) ~ J. Johnson (JJ) (talk) 23:34, 14 March 2017 (UTC)

Papadopoulos 2010 reference

Could someone please fix the reference tagged {{Harvnb|Papadopoulos|2010}} ? It appears as an inline citation in the last line of "VAN method" but there is no corresponding entry in the "Sources" section. SV1XV (talk) 02:33, 14 March 2017 (UTC)

Fixed, @Sv1xv:, and thanks for checking! JerryRussell (talk) 02:52, 14 March 2017 (UTC)

Re recent edits (March 2020)

@Stickyhammer: thank you for attention, as the article really needs it. But I felt there were enough issues with your edits that it will be easier to restart, so I have reverted them. I would suggest copying the article into your sandbox (I can help with that), where you can experiment freely with less disturbance in article space. Particularly regarding citation, where you appeared to be having some difficulty, with your own copy I could give some examples.

I would point out that a significant challenge for this article is that very little work seems to have been done in the last decade or more, and it is likely this topic has been abandoned by mainstream science. That doesn't mean that there is no recent literature on the topic, but that such reports as have been published are likely marginal, have had no confirmation, and there have had no impact. This is why we tend to discourage use of primary sources, because mere publication is not necessarily significant, or even reliable. Preferred are secondary sources where an expert evaluates one or more primary sources. But note that a report in (say) a newspaper of some claim is likely not a suitable secondary source, unless the writer has some competency on the subject.

Please ask if you have any questions. ♦ J. Johnson (JJ) (talk) 20:44, 8 March 2020 (UTC)