Since November of 2015 we have been trying to spread the word about a new method of earthquake prediction (read about it here) and recently (July 2016) we attempted to contact Dr. Jeffrey Love of the USGS, who has made a challenge to the world to predict the timing, location, and magnitude of an earthquake. Unfortunately, what began as a request for help turned into an animus interaction punctuated by a 48 hour flurry of emails that ended with Dr. Love asking us to go further than every solar and planetary physicist before us, and refusing our request for help. The most gut-wrenching part was his claiming that we have never predicted anything, saying so to members of the public who approached him as an authority of the US government, and replying from his USGS email.

If you got here and haven’t read the background on the earthquake forecasting results, please, read it here. This should be in the hands of experts, not me, but the US government refused my offer to give them everything for free and to continue giving input for free. They want nothing to do with this, or us. Someone else needs to catch this hail mary.

This is from an email chain between myself, one of the Observers, and a number of USGS officials. The following USGS employees were copied on the email chain in addition to Jeffrey Love:

My commentary is in red. Apart from removing someone’s name nothing has been changed, including my spelling errors in the emails :)

August 2, 2016 - From an Observer to Jeffrey Love,

“Dear Mr. Love,

Please demonstrate objectivity, lack of prejudice, and an ability to examine and evaluate data, all necessary qualities and skills for your job - which, I believe, is paid for with tax dollars - by responding to Ben Davidson, Dr. Chris Holloman, and Dr. Kongpop U-yen. They have answered your challenge to find a connection between the sun and earthquakes, explained here: http://spaceweathernews.com/challenge/ and here: http://spaceweathernews.com/spf/

If their claims are wrong, explain why. If they deserve further study, then advance science by supporting further studies. That is fair and in the public's interest. That is what is expected of someone in your position.”

-From Love to Observer

“Hi (name removed), thank you for your interest. I haven't challenged anyone to

find a relationship between solar activity and earthquakes. I have,

however, checked published, peer-reviewed reports of such

relationships. The basic difficulty for the folks at spaceweathernews

is that (as far as I know) they haven't actually predicted any

earthquakes before they have occurred, and, then, shown that their

predictions are better than random. When and if they manage to do

that, however, and after they publish their results in a peer-reviewed

journal, I will be happy to have a look.”

August 17, 2016 - From the same Observer to Love

“Hello Mr. Love,

As you're probably aware, your concerns about predictions and publication were addressed in this daily news video made August 8, (https://www.youtube.com/watch?v=aeIdLMyAxfs).  The relevant portion starts at the three minute mark.

As Mr. Davidson notes, he did predict earthquakes, on Twitter, with 37-40% accuracy over the span of a few weeks. In addition, in his Aug. 11 news report, he predicted seismic activity in the southwest Pacific. A few hours later, two earthquakes struck in the southwest Pacific, one at New Caledonia (7.2) and one at Fiji (6.1). He reported on that in his Aug. 12 news report.

The findings of Mr. Davidson, Mr. U-Yen and Mr. Holloman regarding connections between solar activity and earthquakes were published in the journal New Concepts in Global Tectonics, Journal V.3, No. 3, Sept 2015. The papers can be found here and here.

This is an important question of public safety. He has met your criteria. You said (below) if that were to happen, you would be happy to have a look. Will you please have a look and respond to Mr. Davidson (and the public)? If not, why not?”

- Love to Observer

“I look forward to the results being published in a peer-reviewed journal.”

Love hit reply all and this email came to me as the first communication I ever received from Love.

August 18, 2016 - From me to Love

“Hilarious. First, I do not qualify to publish in the journals you consider to be worthy, otherwise NCGT would count. Second, who is going to publish twitter and youtube predictions? Before an earthquake, there is no time - we can’t see the quakes coming more than a few days out and so journals are not an option. Afterwards, I still do not qualify to even open an account to submit manuscripts, and again, who is going to publish twitter and youtube predictions? That is a very unfair standard, and it goes against all logic when all you actually need to do is to look at the timestamps on twitter and youtube. I already responded in this fashion in the Aug.11 update to SpaceWeatherNews.com/Challenge, but I’m happy to repeat it here now.

You are telling people I have not been published and that I’ve never predicted anything. Fact: I have no peer-reviewed works on earthquake predictions. I have two such papers on the sun-quake connection, including a model made by a stats professor which has worked even better since publication ( SpaceWeatherNews.com/SPF ). We have absolutely published on the sun and earth, and we have made numerous successful forecasts of seismicity. Either you knowingly lied to those people or you really didn’t know, which is indicative of the lack of diligence required to discount solar influence over some earthquakes.

I do not care which of those it is, but I imagine the rest of this audience (and mine) care a great deal. You published the paper that is used to crush people like me. You stated the rubric for earthquake prediction. I took on both of those and you won’t acknowledge either one… simply because I don’t have the right advanced degree. Luckily for me, I’m not publishing physics I’m publishing statistics, and Ohio State was kind enough to lend me one of their experts for that. Recall, this whole thing started with “Love has made no mistakes, we just need to look beyond sunspots, geomagnetism, etc, to the solar polar fields”.

For the rest of you, the following are absolutely true facts.

Timeline of Earthquake Challenge:

2013 - Love's paper is touted as the crushing blow to sun-quake proponents. Love's comments on earthquake prediction state rubric for what he considers a prediction. More such comments came in a more-widely-read media article in 2014.

2014 - Davidson, Holloman, U-yen study the full set of SPF data up through 2013 and find a highly significant relationship with earth's largest quakes.

2015 - Two papers published in September in NCGT. Nov 8-25: OLR-based earthquake prediction, with multiple successful predictions of timing/location of M6+ earthquakes, including a magnitude limitation prediction https://twitter.com/TheRealS0s/status/664225661263978496.

2016, July - Davidson accepts Love's challenge, believes that OLR predictions and now, the confirmation of the SPF model into the future, meet comment-worthy status re: Love's paper and prediction rubric.

2016, July - In response to citizens' emails, Love claims Davidson has not been published and has never predicted any earthquakes.

2016, August 8 - Davidson learns of Love’s false claims in private emails. Davidson is not amused. Davidson looks at the Sun, OLR, and earthspots… and sees an opportunity on the horizon.

2016, August 11 - Davidson published a prediction that significant earthquake upticks (which in our community means M6+) will soon strike the area of New Zealand and the islands to the North (Fiji, Vanuatu, New Caledonia, Kermadec, Tonga, Samoa, Norfolk, etc.) https://www.youtube.com/watch?v=CBTKC_asylk&feature=youtu.be&t=1m28s.

2016, August 12 - M7.2 earthquake struck New Caledonia (0126UTC) and hours later a M6.1 struck Fiji. Earthquakes are predictable. Dr. Love is telling people across the country that I've never been published and never predicted any quakes.

History is coming fast; there is still room on the 'right side' for everyone.”

(Since original publication of this email chain, the work has progress much further.)

- Love to me, 11 minutes later

“Ben, actually you can publish in a peer-reviewed journal. As for your predictions, you would need to get them recorded before the earthquakes occur, some place where they can be seen by others and preserved for future scrutiny. Then you would need to establish the statistical significance of the predictions.”

Translation: Dear twitter and youtube, you don’t count as evidence.

- me to Love

“All the predictions are recorded before the earthquakes occur, on VERY public and permanent places, viewed by tens of thousands of people before the earthquakes occur. Is that your way of saying youtube and twitter do not count? As for the statistics, I lack the statistics skills to do it, and wouldn't even know how to describe the parameters. This is a matter of being obvious. The predictions are not many, and the hits are so nearly total that no reasonable person can deny what is happening if they actually take a look.

What journal would any of you consider a good place to public a twitter or youtube prediction and subsequent statistical analysis? Somewhere there isn't a requirement that you be a member of the field? Who will let me publish.

My predictions are not only real and the first of their kind, but they have been preserved for scrutiny in the same ways that celebrities can't seem to remember - there is no deleting these things from the net. As for being 'seen by others' -- ours have actually been viewed by more confirmed unique individuals than any others in history... again, thanks to the internet.

If you have the diligence to watch that Aug11th update on the site you will know what it takes to do an OLR analysis, and then just think about what it took to bust out the lengthy location-forecasting process again and hit the Aug.12th earthquakes we predicted in the morning update from Aug.11th on youtube. I could have emailed you, or made the Aug.11th update days earlier when I first learned that you told people false facts about my work, but I figured it would be better to just show you how it's done.

We were 3 days away from an earthquake watch according to the sun, so I used OLR and coronal hole structure to forecast a seismic uptick for the southwest pacific and published it on youtube, Facebook, and numerous websites at ~6am eastern on the 11th. Early UTC on the 12th the big ones began.

I am hearing excuses not to pay attention. I am seeing ego and stubbornness. Have you even read our papers?

Still a couple things in that previous email that didn't get a response.”

- Love to me, 9 minutes later

“Earthquakes of different sizes occur all the time. If you don't know how to establish the statistical significance of your predictions, then you might read up on such matters in a textbook. This sort of thing is the responsibility of anyone promoting a new hypothesis. You have to show the scientific community that you scrutinized your own results. There are many journals where you might submit well written results. Consider those of the AGU.”

- me to Love

“Jeff, the last time I tried to publish somewhere of that caliber I had a unanimous recommendation for publication through review and then after 4 months they said it didn't fit the issue, and then my account was deleted. I showed the emails and pages online when it happened and that is also preserved for eternity. This was Earthquake Science by Springer, and I'm sure China is now 50 steps ahead of me. Their attempts to keep the rest of the world blind to it are getting a nice boost from you. I am not a physicist, I am not a scientist, I saw a pattern. When I needed a model I went to a statistician and a NASA engineer. Now I need you.

To whom do you think you are speaking? Pretend I'm a 7 year old who was screwing around online and saw this pattern. What would you tell him or her? "Go read a stats book"? I am a member of the public, with a law degree. I am extremely diligent but lack the skills just as a child would in this endeavor. I am not some hot-shot grad student trying to take your job. I'm saying I have done some pretty cool stuff and I can go no further... please help.

Your standards are unreasonable, especially when I've satisfied the spirit of both the sun-quake hypothesis and the prediction element of earthquakes. This technicality in who and where published the forecast does not change the fact that it happened.

I am actually stunned this is where we are... topped off with the insult of a rapid trigger finger on the send button letting me know you clicked no links, did no homework, checked no facts, and just will continue to tell yourself, and others, that I haven't done the things that they see me do all the time.”

- Love to me

“Ben, if you haven't yet shown your results to be statistically significant, then I think you should refrain from further promoting them until you do.”

- me to Love

“PS - this is about the papers, remember... that's what this challenge is about. I didn't get into the prediction thing until you told people I never did it. SO I did it again. It takes too long to do these, and as you said, nobody has published our social media predictions... but then again, that's not really what this is about. It's about the peer-reviewed papers. Please don't deflect back to the predictions again; it is not the crux of this issue although it IS the more impressive, by far.”

So I clearly attempted to focus the conversation to the published work and not the predictions.

- Love to me

“No deflection is being made. For your predictions to gather serious attention, you will need to both show statistical significance and publish in a peer-reviewed journal. I'm trying to help you by emphasizing those facts.

So you say, now, that you can show that the results are significant? Can you tell me about your null hypothesis and what statistical test you used?

Also what p-level did you find?”

Right back to the predictions…

August 19, 2016 - me to Love

“1) Location Predictions/Forecasts: Every single time the public has EVER gotten excited about a prediction... it failed. I'm thinking of that Simon guy who called out the North Atlantic to rumble a few years ago. I've actually done it. I'll work on the stats part of prediction... but honestly... I can take this to the media now and your arguments will sound trivial and small.

2) THE ACTUAL CHALLENGE - The SPF Trigger: As for the actual published peer-reviewed paper, Dr. Chris Holloman of The Ohio State University did the statistical analysis, and if you can pick out the name of a 'method' in his 1000+ lines of R code be my guest. The p value was ~0.000015 or something in that range.

The null hypothesis was that peaks in magnetism and magnetic reversals of the SPF had no correlation, temporally, with the largest earthquakes- that there would be the average share of earthquakes in those days. Example: if we said 10%, 30%, 50% of the days were significant, we should see the same share (10, 30, 50%) number of M8+ earthquakes in that time.

This information is on the challenge page and in the peer-reviewed paper.

After all this... you still didn't read the paper. Do you need the link again? You can find both of them at New Concepts in Global Tectonics repository from Sept. 2015, or you can find them (and the subsequent research notes) at SpaceWeatherNews.com/SPF 

Here is a link to the Foundational Paper that made it through peer review twice. This is what the Chinese government is using to begin their efforts.

Ben

PS- if you want, I'll send you the video where I said Simon should be looking at Kamchatka instead... and then the link to the USGS page for the 8.4 that struck later that week in May of 2013.”

- Love to me

“Statistical significance applies to predictions, not to retrospective fits of models. Again, i encourage you to submit your results to a peer-reviewed journal.”

Again, the predictions. Which is why we focused there since this paper was originally published in August of 2016.

- me to Love

“The following can all be found on the challenge page OR in the paper (from the peer-reviewed journal) itself:

The statistical significance is based on the chances that so many of the earthquakes would fall so close to the important events, within what we called, "significant windows". According to the null hypothesis, we should have seen about 40% of the M8 earthquakes since 1976 fall within those windows, instead it was nearly 80%. That is like having a coin rigged to come up heads 60% of time, and yet after 33 coin flips you got tails 80% of the time. The p value is the chances of that result, given the n and other relevant items. If you are suggesting that this professor from Ohio State doesn't know what he is doing then you can take that up with him. Every statistician who has read the paper understood that concept, had no problem with the method, and many have kept up with the progress. The chances of a 50/50 coin coming up 80% one-way after 33 flips is what? Now rig that coin against you 60/40.

The model itself is based on peaks of magnetism and magnetic reversals - that is all.

If there is no correlation the earthquakes should be randomly scattered over time, with no relationship to the solar events of note. Each polar field peaks about once per year, and every 11 years we get our reversals. These events are separated by months to years. The average number of days between these events is 110 days, which means we should see an average days-from-solar-event for each quake of about 27.5 days, with them being randomly scattered from 0 to about 55 days away from the solar event.

In reality, the average is 14 days. Half of the earthquakes should have been more than 27.5 days from the solar event. Only 6% were more than 27.5 days away.

If no correlation, 4 lucky quakes among the list of M8+ would be within 7 days of a significant solar event, and 8 would be within 14 days.

In reality, 14 of them were within 7 days, 20 were within 10 days.

In 2015 and 2016, as of the July update, there had been 10 M7.5 events (not enough M8’s). 8 of 10 quakes were within 11 days of significant solar events, and half of those were within 3 days of the solar event - remember, the average should still be 27.5 days.

Some FYI on the reviews process...

By the way, both Earthquake Science and NCGT gave recommendations for publication. Yes, the issue of ES was not a fit for our work, according to them, but the review process at NCGT was vastly more difficult. With Springer, the reviewers asked that I use "earthquake trigger" rather than "earthquake precursor" which I used throughout the work, initially. Another commented that a cited paper wasn't used. I changed the language, informed the reviewer where the citation could truly be found in the manuscript (appeared to be an English/Chinese language barrier), and then they unanimously said it should be published. That was it. Then I went and I got my butt kicked by NCGT, where the rigor and difficulty of the process was 10-fold. Amazingly, they ended up being the ones to publish it. So, two successful peer reviews, an administrative decision, and finally, publication of the papers.

Please give me more than 30 seconds before you email back so I have time to research the veracity of this "you can't run statistical significance on a retrospective dependency analysis between two data sets" claim you are making.

Peanut gallery, hope all is well.”

- Love to me

“Statistical significance is established from predictions, where you predict the earthquakes before they occur.”

- me to Love

“You have your challenges confused. This began as a response to your 2013 paper, which is used by 99.99% of people to tell me why SPF cannot possibly be related to quakes - even though your work was not focused on those things. You made no mistakes in your analysis, and in fact I said the same thing in 2012 - that there are some interesting coincidences, but no long-term relationship between SSN, flaring, CME impact, Dst, induced current, riometer spikes or solar wind speed. I actually went a bit further than you did.

The challenge began as you being the only person who might be able to convince anyone this relationship was valid, since you are the primary person used to say people like me are grasping at invisible straws. Then you told a bunch of people I've never predicted anything, and so I added in the prediction stuff at your inferred request. The paper has nothing to do with the twitter and youtube predictions.

Retrospective analysis can't be used? In fact, it is the entirely of statistical significance measures in the history of earthquake science, since there has never been any successful predictions published. Since I assume you didn't mean that retrospective analyses cannot be used to show statistical significance of anything, I am forced to think that you are confusing that challenge with the prediction challenge, the rubric for which is best found in your comments to media in that 2014 article about electromagnetic precursors to earthquakes.

These are separate issues. If you want to get technical, we published the model, which itself forecasts that the big quakes will happen along with the specific solar events... even though the 4week delay in data delivery prohibits predicting specific events, when the data came out and eventually showed that the big quakes were STILL hitting those marks, that is a prediction-hit in and of itself.”

Clearly… the focus is once-more put on the PR-published work and not on the predictions.

- Love to me

“Again, if you think you can predict earthquakes, then I say predict them. Then, after you predict them, show that your results are statistically significant, and, then, after you scrutinize your own results (and even try as hard as you can to disprove them), after all that, write it up, submit it to a peer-reviewed journal, and then you will get the attention of the scientific community. As you note, this has never been done, and this is the challenge for earthquake prediction. Retrospective "predictions" are not how you show your method to be useful. It is normal that scientists scrutinize their own results. It is not normal for scientists to ask others to do it for them.”

- me to Love

“Again, you have your challenges confused. The prediction side of things is secondary, wasn't part of how this all began, and was actually introduced by you when you said I didn't predict quakes.

I am not talking about prediction and SPF together - Stanford has a 4-week delay on the data so it's impossible. My entire point about quakes is that you can say I haven't been peer reviewed, but you might also mention that I've 3 for 7 on twitter and 1 for 1 on youtube, with the 1 being on your command.

I am talking about the analysis of space weather, like you did in 2013. If retrospective analysis cannot be used then what is data comparison all about? This began as me responding to your 2013 paper about a relationship between the sun and earthquakes... the prediction stuff is a side-note we're having that you can't seem to get out of your head.

Challenge 1: Show a relationship between the sun and earthquakes.

Challenge 2: Meet Love's earthquake prediction rubric.

I'll give you that I have not been peer-reviewed on #2, but then again, this whole Davidson/Love thing started as wholly #1. You brought up #2, not me. I want to focus on #1, and you cannot tell me retrospective analysis cannot be done for data set comparison to get a statistical significance of the existence of the relationship.

To suggest a retrospective analysis cannot be performed for SPF makes me wonder why you wasted your time on sunspots and geomagnetic storms in the first place. None of those works you reviewed contained ANY predictions, other than the continuation of the model itself (which all failed, I agree, until mine) and indeed were all retrospective analyses of space weather indices.

In the time you have dedicated to these emails, you could have read all 4 papers, watched all 4 videos on the challenge page, copied/pasted the SPF data into a spreadsheet and done this yourself. That last part literally takes 3 minutes.”

- Love to me

“It is perfectly valid to reject a hypothesis on the basis of retrospective data analysis. It is not, however, valid, the validate a hypothesis on retrospective analysis. For that you need to prospectively make predictions. This is how objectivity is attained. I analyzed sunspots, solar-wind, and geomagnetic data because those were some of the data sets used by other researchers (and reported in journals), and I found that their results didn't hold up. If you find a new method, based on a new hypothesis, then it still remains for you to predict earthquakes and show that you can do better than random. This is explained, yes, in my paper.”

- me to Love

“So, when our model performs better into the future, like it has so far, what does that mean? You literally just made every point I was trying to make in those last emails. This has ALL been about challenge #1. Why not include #2? I didn't feel ready, and you agree. However I DID have four M6+ hits on time/location in a two-week span in November (that was all I could handle given the enormity of the analysis of OLR, GEC, etc) and upon your telling people I never did it, I predicted the next M7 three days later just last week. I never intended for the prediction aspect to be part of our discussion - but you insisted.

Now, back to the real issue, the one I AM ready for. The confirmation of the model. I wanted you to be a part of it. It is as simple as publishing the statistics in my previous email, and this isn't as hard as making the model - it's just seeing how absurdly close, within days, all of the recent quakes are and putting new numbers in the equation. It is literally GUARANTEED to be better than before - go back and re-read that section of the previous email.

Now we come down to what you may recall (or maybe not, since you didn't do much homework on this interaction or my work) is the entire purpose of reaching out to you in the first place. I did so with praise, both for your career and your 2013 work, but Dr. Love, I have also been having nightmares about your name for the last 2 years- in the figurative sense. Your work has been used as justification for mine to be ignored and ridiculed. I have a way for THE authority on this matter to confirm what so many of us can plainly see to be true, and in doing so, needs not ever admit wrongdoing or error. The error is wholly on the history of publishers in the solar/quake realm for never looking in the right place, and you rightly quashed their previous efforts. You could be the one to stomp down the failures AND hold up the holy grail.

Can you guess what I wanted you to do with me when I initially reached out to you?”

- Love to me

“Rejection of the null hypothesis (which is, technically, not the same as confirmation of the model) comes with prediction. The two are not and cannot be separate. And, again, since this is your model, one that you are promoting, it is up to you to test it.”

- me to Love

“So the model working into the future does not count as a prediction? We predicted that the model would work into the future. It is.

If you need me to use the SPF analysis to predict an actual quake like I do with other factors on twitter and youtube (yes, there are separate factors for these challenges) then I will need Stanford to give us the data in real time. I think that is unreasonable. The model itself is the prediction... that the model will work into the future. It works.

"it is up to you to test it." -- It IS tested. It passed, above the curve. That is where we stand now, with your request being moot, answered, done, etc. I do not know how else to say this.”

- Love to me

“Great! Now publish it in a peer-reviewed journal!”

- Me to Love

“Yes, the plan is to continue publishing again and again... but since it is so blatantly obvious I was thinking maybe you'd like to be a part of it. Something you might recall, I was asking for some help - it is a function of knowing my limits. Your attitude towards this entire matter has brought about the tone of this interaction.

Now... to see if AGU or Atmos Sol-Ter Phys will let me do this on my own (another reason I wanted some help from someone in the field).

This has been a most-unfortunate interaction, one that should have ended up with you getting all the credit for the understanding. Since I can tell this is what you've been itching to hear all along, we're done here.”

- Love to me

“No need to share the credit with me. I look forward to seeing your results in published form!”

.end of emails.

UPDATE: DECEMBER 27, 2016

The last 5 or 6 emails really disappointed me and many of those who have read them in the last few months. This has come WAY further than an amateur should be able to take it- in two different vectors of earthquake science. Then we asked for help from the man who put his flag in the ground in this soil. He refused to help, refuses to acknowledge facts and deflected to have others review the work instead, went most of the way through that chain without reading the peer-reviewed work, never went to the challenge page and heard what was actually going on, and then asked us (even after saying ‘we’ve gone as far as we can go’) to go further than any solar or planetary physicist in history… further than any of those he was willing to judge in his 2013 paper.

If you understand the “7-yr-old discovers pattern” example then you realize how preposterous this scenario truly is. Despite the evidence, when it comes to saving lives by predicting earthquakes, we, the people, are very-much on our own.

In October and November of 2016, we sought to determine the viability of such a requested published work in a major journal. Overwhelmingly, and without fail, there was no journal willing to even consider such publications. Love’s demand is categorically impossible.

So what now? I can’t pretend I’m not stoked at the idea that Dr. Love challenged us to do more than everyone else he previously reviewed, despite our lack of skills and requests for help with doing just that very thing. I can’t pretend I am not entertained by what appears to be a lack of attention and diligence on the part of Dr. Love. The world can never pretend that he has not laid out specific criterion now to meet his standard of review. Keep up with how things are shaking out at QuakeWatch.net

-

Space Weather News is an Albuquerque company devoted to reporting on space weather, earth weather, earthquakes, and science, and also to original research in solar, terrestrial, and solar-terrestrial physics. It hosts the most-watched daily space weather news program on the internet and maintains a number of resources: SpaceWeatherNews.com, EarthChanges.org, Suspicious0bservers.org, ObservatoryProject.com, QuakeWatch.net and MagneticReversal.org.

Contact

Kat@ObservatoryProject.com

Ben@ObservatoryProject.com