Hacker News Reader

Comment by matthoiland | Hacker News Reader

>RF Seismograph showed an increase in noise on 80 meters some 13 hours beforehand, as well as some propagation changes on 40 and 30 meters — low before the quake and increasing in its wake.

Hours. Hours beforehand. Having that much preparation would be life-saving. Especially for us living near the Cascadia Subduction Zone.

View on HN

Keep in mind that this report has come about with zero peer reviewing (I think it's a little strange it's been published by the ARRL to be honest), and another thing that is missing is that it needs to be correlated with solar activity, operator activity, local noise, and even thunderstorm activity, among other influences, before one can validate a change in HF propagation is a precursor to earthquakes.

All these things sound like stuff which can be controlled for.

Absolutely, but it just needs to be done.

Because there are also a million other correlations with the earthquakes which are just coincidences.

-Which is all the reason I need to bring up Tyler Vigen’s amazing spurious correlations:

https://www.tylervigen.com/spurious-correlations

That's hilarious. Would be cooler if they could limit it to stuff with the same units though!

> Keep in mind that this report has come about with zero peer reviewing

This report, maybe, but even the article includes links to previous ones which have.

Here’s one, but if you click through to the project’s website you’ll find more.

http://www.ep.sci.hokudai.ac.jp/~heki/pdf/Scientific_America...

I've always assumed that piezoelectricity would occur in the crust before and during an earthquake given the prevalence of quartz - it's my go to theory for the now undeniable sightings of earthquake lights.

There are other things that can cause E skip to stop working (solar activity is the big one) so while it is good to say "hey propagation just changed, what's up?" it isn't reasonable to say "Earthquake inbound!"

That said, there is interesting work in geology to more accurately characterize this mechanism. The last paper I saw was reference in Science News indicated that pressure on granite (which has a peizo-electric effect) might be the underlying mechanism. I haven't seen any follow up on that hypothesis though (proving or disproving). It would seem pretty straight forward for the seismograph sensor makers to add an E field strength meter to observe this effect if it occurs.

The creator of FiveThiryEight, Nate Silver [1], had a great chapter in his probability book about the global efforts to detect earthquakes and there have been tons of research over the last century (including radio). With many people claiming to have found solutions only for them to quickly fail.

With current best-effort science it’s basically impossible to detect them ahead of time, particularly with any relevant geographic accuracy. At most only a probability figures within a 10-100yr+ range over a very broad area seems to be the best they can do. Fault lines can encompass hundreds of towns and cities.

Even if they detected the Cali quake before hand from Vancouver there is no way they could pin point that in any reasonable way to evacuate particular towns AFAIK.

1: https://books.google.ca/books?id=SI-VqAT4_hYC&lpg=PP1&pg=PA1...

RF Seismograph showed an increase in noise on 80 meters some 13 hours beforehand, as well as some propagation changes on 40 and 30 meters — low before the quake and increasing in its wake

Does this mean my cat is tuned in to ham frequencies? She was not right in the hours before the 7.1, and once the shaking stopped she was back to her normal self.

I had a cat growing up and he always acted very strange before significant earthquakes.

Did your cat ever act very strange at other times?

Did significant earthquakes ever occur when your cat was acting normally?

The classic parody textbook "Science Made Stupid" theorized that earthquakes are triggered by vibrations from agitated animals.

(http://www.chrispennello.com/tweller/ , page 38)

How often does this RF Seismograph show disturbances when no earthquake follows?

Only if the test has high specificity. If this RF noise happens frequently when there is no earthquake, it is of little value.

Little PR value, to be sure- but still fairly valuable if someone can identify the causal factor and more specifically measure that.

The YouTuber Suspicious0bservers somewhat predicted the Quakes via Space Weather: https://www.youtube.com/watch?v=ISfueR4vAhQ. His channel and research are interesting.

The U.S. Geological Survey does issue notifications when they detect significant seismic activity that's a potential precursor to larger events. But as many people have already mentioned, there still is no known model for predicting earthquakes.

With inexpensive, distributed RF software defined radios, we could rapidly collect data and use machine learning to see if a pattern presents itself based on changes in HF propagation.

For a similar example, check out https://www.lightningmaps.org/?lang=en. They use inexpensive, freely colocated time of arrival RF receivers to triangulate lightning strikes across the world [1]. It operates on a shoestring budget as a hobbyist project.

There might be no signal to be found, but it seems from a cursory glance that to investigate would be a low cost exercise.

[1] https://i.imgur.com/dUxnTsR.png

I don't think it has been consistent as far as observations go, so its usefulness in prediction seems unknown at this point.

Reply on news.ycombinator.com