Talk:VAN method

Work update ( section "Predictive success" )

The following content is suggested to be added to the third paragraph of the section: "Predictive success":

On the other hand, the Section "EARTHQUAKE PRECURSORS AND PREDICTION" of "Encyclopedia of Solid Earth Geophysics: part of "Encyclopedia of Earth Sciences Series" (Springer 2011) edited by Harsh K. Gupta, ends as follows (just before its summary): "it has recently been shown that by analyzing time-series in a newly introduced time domain "natural time", the approach to the critical state can be clearly identified [Sarlis et al. 2008]. This way, they appear to have succeeded in shortening the lead-time of VAN prediction to only a few days [Uyeda and Kamogawa 2008]. This means, seismic data may play an amazing role in short term precursor when combined with SES data". Furthermore, in 2018, the statistical significance of the VAN method was revisited by employing modern techniques, such as event coincidence analysis (ECA)[Donges et al 2016] and receiver operating characteristic (ROC) [Fawcett 2006], which revealed that SES exhibit precursory information far beyond chance.[Sarlis 2018]

The above contains the exact excerpt from the encyclopedia. I feel should be reproduced in the article, as it answers an open question. The rest of the citations are:

• Donges, J.F.; Schleussner, C.-F.; Siegmund, J.F.; Donner, R.V. (2016), "Event coincidence analysis for quantifying statistical interrelationships between event time series", The European Physical Journal Special Topics, 225 (3): 471–487, arXiv:1508.03534, doi:10.1140/epjst/e2015-50233-y, ISSN 1951-6401
• Fawcett, Tom (2006-06-01), "An introduction to ROC analysis", Pattern Recognition Letters, ROC Analysis in Pattern Recognition, 27 (8): 861–874, doi:10.1016/j.patrec.2005.10.010, ISSN 0167-8655
• Sarlis, Nicholas V. (2018), "Statistical Significance of Earth's Electric and Magnetic Field Variations Preceding Earthquakes in Greece and Japan Revisited", Entropy, 20 (8): 561, Bibcode:2018Entrp..20..561S, doi:10.3390/e20080561

If no objection is expressed, I will update the article.--EyeCont (talk) 08:36, 27 April 2020 (UTC)

• improvement by Mikenorton   ManosHacker ^talk 04:35, 1 June 2020 (UTC)
• improvemet of citations by Mikenorton   ManosHacker ^talk 04:35, 1 June 2020 (UTC)
• sum replacing the lengthy quoted text and hiding of the direct visibility of the "Encyclopedia of Solid Earth Geophysics" as the source, by Mikenorton   ManosHacker ^talk 04:46, 1 June 2020 (UTC)

Literature update ( section "Electromagnetic compatibility issues" )

A literature update is missing at the end of the section:

In 2003, modern methods of statistical physics, i.e., detrended fluctuation analysis (DFA), multifractal DFA and wavelet transform revealed that SES are clearly distinguished from those produced by human sources, since the former signals exhibit very strong long range correlations, while the latter signals do not.[Varotsos, Sarlis and Skordas 2003a][Varotsos, Sarlis and Skordas 2003b] A work published in 2020[Christopoulos, Skordas and Sarlis 2020] examined the statistical significance of the minima of the fluctuations of the order parameter κ1 of seismicity by event coincidence analysis as a possible precursor to strong earthquakes in both regional and global level. The results show that these minima are indeed statistically significant earthquake precursors. In particular, in the regional studies the time lag was found to be fully compatible with the finding[Varotsos et al 2013] that these mimima are simultaneous with the initiation of SES activities, thus the distinction of the latter precursory signals from those produced by human sources is evident.

• Varotsos, P. A.; Sarlis, N. V.; Skordas, E. S. (2003a), "Long-range correlations in the electric signals that precede rupture: Further investigations", Physical Review E, 67 (2): 021109, Bibcode:2003PhRvE..67b1109V, doi:10.1103/PhysRevE.67.021109, PMID 12636655
• Varotsos, P. A.; Sarlis, N. V.; Skordas, E. S. (2003b), "Attempt to distinguish electric signals of a dichotomous nature", Physical Review E, 68 (3): 031106, Bibcode:2003PhRvE..68c1106V, doi:10.1103/PhysRevE.68.031106, PMID 14524749
• Christopoulos, Stavros-Richard G.; Skordas, Efthimios S.; Sarlis, Nicholas V. (2020/1), "On the Statistical Significance of the Variability Minima of the Order Parameter of Seismicity by Means of Event Coincidence Analysis", Applied Sciences, 10 (2): 662, doi:10.3390/app10020662
• Varotsos, P. A.; Sarlis, N. V.; Skordas, E. S.; Lazaridou, M. S. (2013-03-18), "Seismic Electric Signals: An additional fact showing their physical interconnection with seismicity", Tectonophysics, 589: 116–125, Bibcode:2013Tectp.589..116V, doi:10.1016/j.tecto.2012.12.020

If no objection is expressed, I will update the article.--EyeCont (talk) 09:20, 27 April 2020 (UTC)

Restore accuracy ( section "Public policy" )

Just after the "the VAN group issued a series of telegrams in the 1980s" follows this: , warning of impending earthquakes that did not occur, or did not occur within the parameters listed in the telegrams. During the same time frame, the technique also missed major earthquakes.[Hamada 1993], which is cited with Hamada 1993. This is inaccurate and should be replaced with the exact text of Hamada 1993 in order to restore accuracy, as follows: During the same time frame, the technique also missed major earthquakes, in the sense that[Hamada 1993] "for earthquakes with Mb≥5.0, the ratio of the predicted to the total number of earthquakes is 6/12 (50%) and the success rate of the prediction is also 6/12 (50%) with the probability gain of a factor of 4. With a confidence level of 99.8%, the possibility of this success rate being explained by a random model of earthquake occurence taking into account the regional factor which includes high seismicity in the prediction area, can be rejected". This study concludes that "the statistical examination of the SES predictions proved high rates of success prediction and predicted events with high probability gain. This suggests a physical connection between SES and subsequent earthquakes, at least for an event of magnitude of Ms≥5".[Hamada 1993]

• Hamada, K. (August 1993), "Statistical evaluation of the SES predictions issued in Greece: alarm and success rates", Tectonophysics, 224 (1–3): 203–210, Bibcode:1993Tectp.224..203H, doi:10.1016/0040-1951(93)90073-S

If no objection is expressed, I will update the article.--EyeCont (talk) 10:42, 27 April 2020 (UTC)

Citation still needed ( section: "Public policy" )

Regarding "Major opponents of VAN were the Greek seismologists Vassilis Papazachos and G. Stavrakakis. The debate between Papazachos and the VAN team has repeatedly caused public attention in their home country Greece and has been extensively discussed in the Greek media.[citation needed]", although a citation for this excerpt has been asked for, not any citation has been provided for a long time. In order to retain this excerpt, please provide a citation.--EyeCont (talk) 10:59, 27 April 2020 (UTC)

Recent changes

I have not looked in detail at the proposed changes that have now been enacted. I've just been busy elsewhere and this was quite a lot to take in. I will now try to find some time to consider these changes and probably add some more sources that are critical of the way that the VAN method is currently implemented, because it now reads like all issues have been answered and everyone's happy with the method, which is definitely not true. As I said elsewhere, my take on this (and it's only a guess and impossible to prove) is that most seismologists involved in the field of earthquake forecasting/prediction are just ignoring the VAN method. Best of all would be examples of reviews by seismologists not previously involved in the debate regarding its applicability. Mikenorton (talk) 11:57, 13 May 2020 (UTC)

@Mikenorton:Top seismologists are using the seismological analysis part of the updated VAN method extensively nowdays and apply it to earthquake nowcasting.   ManosHacker ^talk 10:48, 23 May 2020 (UTC)

@ManosHacker:I presume that you're referring to papers by John Rundle, Dan Turcotte and others applying "natural time" analysis to minor earthquakes as a way of estimating the current state of stress and strain accumulation. This has nothing to do with the VAN method as such, which is based on SES. I've just looked back through Google Scholar results from 2017 to now for the "VAN method" and I found that nearly half were published by members of the VAN group or people that have previously published papers with them (14 out of 33), 6 made use of some aspect of the method (generally "natural time", including one in medicine) or looked at SES in general but did not apply the method to prediction, 10 mentioned it either in passing or in one of the references used, 2 provided support for the idea (1 of which had nothing new to add) and 1 was critical. Nothing there to suggest that there is any significant take-up of the method. This paper by Hayakawa is the only one to look positively at the applicability of the method by someone outside the group, including his own post-hoc "prediction" of the 2011 Tohoku earthquake. The critical one is this paper by Helman, which I've mentioned elsewhere. Mikenorton (talk) 18:05, 23 May 2020 (UTC)

@Mikenorton: An earthquake prediction method should determine of course the epicentral area, the magnitude and the time window of the impending mainshock. Varotsos' book "The Physics of Seismic Electric Signals, TerraPub, Tokyo (2005)", the basic text for the VAN method, explains that the epicentral area and the magnitude are determined from the SES. In addition, the initiation of a SES activity marks the time at which the natural time analysis of the subsequent seismicity should start, in order to determine when the mainshock is approached. Hence, there are two cornerstones of the VAN method, namely the SES and natural time, both introduced by the VAN research group in the beginning of the 1980s and 2000s, respectively. Thus, it is not correct to say nowdays that VAN method is solely based on SES. The eminent scientists Dan Turcotte, John Rundle & coworkers, upon introducing in 2016 their challenging method "nowcasting earthquakes" using natural time, explain clearly that "nowcasting" answers the question of the current seismic hazard state of the region. It is distinct from forecasting and in fact should be a prerequisite to forecasting, the future state of the system. Finally, in going through ISI Web of Science after 2017, one finds several tens of papers from various countries citing favourably the VAN work, and one was critical as you say. This criticism, however, has been shown (https://www.mdpi.com/1099-4300/22/5/583/htm) that does not stand.--EyeCont (talk) 14:48, 25 May 2020 (UTC)

@Mikenorton: According to the bibliography these recent changes are scientifically correct. Looking at Helman's whole work I do not see this paper of his as an appropriate review but rather as another example of misunderstanding of the physics of VAN method. Hayakawa is a top unrelated seismologist who does a review on VAN method. A fact (rather than scientific debates in equilibrium in questions and answers) missing from the lead section is that VAN method is being systematically applied since 1981 in Greece and is being funded by the state all along. Next is to summarize VAN method for earthquake prediction article.   ManosHacker ^talk 09:05, 30 May 2020 (UTC)

They are correct as far as the VAN group is concerned. You can't just dismiss Helman's criticism on the basis of the VAN group saying that they are invalid, because of course that's what they claim. I do not have the right background to be able to judge and I rather doubt that you do. Hayakawa is a good geophysicist, but I don't see how you can rate him as a "top seismologist". Who are these other "top seismologists" that are applying this prediction method? Where are the successful predictions (and I mean in advance rather than after the fact) as judged by people not in the VAN group? Why are no countries making use of this method to warn populations? Why after 40 years is this method still not being routinely used for this purpose? There are a lot of questions that are not answered as the article stands at the moment. Mikenorton (talk) 10:46, 30 May 2020 (UTC)

@Mikenorton: On the point you raised "where are the successful predictions...as judged by people not in the VAN goup?" there are examples that are unquestionable:

1. Greece: the strongest for decades, MW 6.9 earthquake occurred on February 14, 2008 in Southwestern Greece, and the warning to the population was issued four days before, on 10 February 2008. See the two papers in EOS by Uyeda & Kamogawa in 2008 & 2010.
2. Japan: the mega MW 9.0 Tohoku earthquake occurred on March 11, 2011, before which Japanese and Chinese scientists, not related with VAN group, analysed the magnetic field data at permanent Japanese stations lying 130 Km far from the epicenter and found that anomalous magnetic field variations appeared from 4 to 10 January 2011, showing the generation of strong SES activity almost 2 months before the mega event (see all the the relevant literature in"MW9 Tohoku earthquake in 2011 in Japan: precursors uncovered by natural time analysis").
3. Mexico's largest earthquake in more than a century, the M 8.2 Chiapas earthquake on September 7, 2017: Mexican researchers, not related with VAN, by using natural time analysis had published in Physica A - well before this EQ- that an extreme fluctuation (large earthquake) had the highest probability to occur in southern Mexico in Chiapas area, where the large earthquake finally occurred. This exact area was highly unexpected by the seismologists to give an earthquake, as being aseismic. (additionally explained here)

As far as the changes to the current article are concerned, they can be addressed one by one above, without the need to be an expert to see if the content is verifiable or not. I prefer building of consensus on these in a simple way. The questions you are raising should be converted to facts and put to the article in the lead section. The debates on topics belong to the sections that follow.   ManosHacker ^talk 11:20, 31 May 2020 (UTC)