Can earthquakes be predicted?

Total Pageviews

Search This Blog

Friday, March 18, 2011

Evaluation of earthquake prediction theories(other than plate tectonics) by Science

hi
At the out set ,let me make it clear that, these are my views and I do not intend to offend any branch of Science or any body.
There are three reasonns ,why ,Other theories (specially tidal pull theory ) could not score above average
1)The averages of earthquakes occuring in a year are calculated ,on the basis of total number of quakes occuring ,incluing pre shocks and after shocks. Where as while evaluating ,only one hit per date is taken in to account,though ,there are several quakes on given date.
Considering all quakes occuring ,while calculating odds and taking in to account only one hit per day,is not justified.Either ,we take them all ,or we take only one quake per day,equally in both the cases(ie while calculating odds and while evaluating)
2) Whenever Scientists/Seismologist have evaluated the tidal pull theory ,they have only taken in to account Moon Phases(only one indicater).Very few must have taken in to account both ,phases of Moon as well as Distance of Moon from Earth ie Perigee. Thus ,only partial theory is evaluated ,which can not excced Odds ,calcualted as 1) above.
Recently ,one my friend Roger have tried to complile several indicaters simultenously .
3) Third reason ,I do not want to disclose,as it may offend science faternity
regards
Amit

44 comments:

srikanth said...

Hi Amit.

I see in your posts that you claim to have predicted the earthquake dates.

I would want to you create a new post and predict the major earthquakes in the next month.
Mentioning the location would be great.

I am a hindu by birth. A Practicing brahmin, a believer in god. My Grandfather is an astrologer. But Astrology has serious limitations, according to our scriptures themselves! There is a popular story which hints at astrology being cursed by goddess Parvathi, which made it only half-true. This story/fact makes me even more skeptical about our astrology. Even Science has limitations, because science is nothing but human observation.

Anyway, let us test your claims of predictions once and for all. please make your predictions now! Time shall verify. :)

AMIT said...

Srikanth
pl see my blog or web page

http://earthquakepredictionbyamitdave.blogspot.com/

http://earthquakeprediction.webs.com/ and go to calender
Amit

Random Thoughts Tracy said...

Did you see this?
http://www.nbcbayarea.com/news/local/Water-Spout-Spotted-Off-Ocean-Beach-118262044.html

Anonymous said...

Did you see this tornado in San Franscico this AM? http://www.nbcbayarea.com/news/local/Water-Spout-Spotted-Off-Ocean-Beach-118262044.html

tiffany34 said...

It's amazing how bad things have gotten. Just imagine. Anyone who lives near a fault line should really think about getting some supplies in case of an emergency. There are lots of places like www.buyearthquakekit.com that you can pick up some simple stuff that might just make the difference. Thoughts and prayers to those in Japan.

Roger Hunter said...

Amit;

Here are the results for the elements I have examined. Moon phase, perigee have 1 three day window each. Declination has 1 three day for each of 3 positions (maximum, zero, minimum). Planet stations have 2 three day windows each, one 15 days before, one 15 days after (centered on 15)

7670 Days.
302 Quakes.
Odds on a quake: .113

259 New moons.
26 New moon hits.
260 Full moons.
27 Full moon hits.

279 Perigees.
35 Perigee hits.

281 Positive declinations.
30 Positive declination hits.
562 Zero declinations.
61 Zero declination hits.
280 Negative declinations.
37 Negative declination hits.

14 Venus direct.
13 Venus retrograde.
10 Mars direct.
10 Mars retrograde.
20 Jupiter direct.
19 Jupiter retrograde.
20 Saturn direct.
20 Saturn retrograde.
126 total stations.
2173 total windows.

53 Moon hits.
35 Perigee hits.
128 Declination hits.
31 Station hits.
247 Total hits.
245.6 Expected hits.

Exactly as expected.

Roger

AMIT said...

Roger

One clarification please
You said total 302 quakes (I presume 7+),than how come total hits 247? and expected 245.6?
Odds on quake is 0.113,does that mean if I take a random date there are 11.3% chances of hitting the quake?
Total stationed are 126 than how come 2173 windows? Are those total windows for all indicators?

Roger ,I give 30 odd dates in a year after considering all indicators ie 90 days window.Your evaluation says total 2173 windows (in 20 years)
There is something that s missing.

Now seven indicators I generally use are(in order of importance)
1)Major planets changing the direction
2)Moon at perigee
3)Moon Declinations
4)Phases of Moon 0,90and 180degrees with Sun
5)Moon aspects with major planets
6)Planets distances from earth
7)earth distance from Sun In rare cases I use
Sun and Moon at Same declinations
Mon at Equator as indicators.

I presume You have considered most of them
Moreover ,you can not take random three day window on both side (+ or - 15 days ) of stationed planet If you have taken random 3 day window on both side of stationed planet ,it is not correct. There is a process of selecting date from 15 days on both side

regards
Amit

AMIT said...

Roger
one more thing
Total days 7670
Total window 2173 ie 28% of days

Total quakes 302
This is no software ,but if I take 28% of 302(ie total quakes) ,it comes 85 quakes.(by simple rule of three).Where as software says expected hit as 245.6
Could ou please elaborate,why such vast difference?
In that case,as against 85 ,the hits are 247 may be three times higher
regards
Amit

Roger Hunter said...

Amit;

> One clarification please
You said total 302 quakes (I presume 7+),than how come total hits 247? and expected 245.6?

Why not? The hits are counted individually so one quake may saisfy more than one window.

> Odds on quake is 0.113,does that mean if I take a random date there are 11.3% chances of hitting the quake?

Correct.

> Total stationed are 126 than how come 2173 windows? Are those total windows for all indicators?

Yes.

> Roger ,I give 30 odd dates in a year after considering all indicators ie 90 days window.Your evaluation says total 2173 windows (in 20 years)
There is something that s missing.

You have multiple windows for some indicators.

> Now seven indicators I generally use are(in order of importance)
1)Major planets changing the direction

Have that.

> 2)Moon at perigee

Have that

> 3)Moon Declinations

Have that (3 of them)

> 4)Phases of Moon 0,90and 180degrees with Sun

Have only new and full moons. (0 and 180). What about 270?

> 5)Moon aspects with major planets

No.

> 6)Planets distances from earth

No.

> 7)earth distance from Sun In rare cases I use
Sun and Moon at Same declinations

No.

> Moon at Equator as indicators.

Isn't that zero declination? Have that.

> Moreover ,you can not take random three day window on both side (+ or - 15 days ) of stationed planet If you have taken random 3 day window on both side of stationed planet ,it is not correct. There is a process of selecting date from 15 days on both side

Please explain. I take 3 day windows centered on 15 days before and after station.

Roger

Roger Hunter said...

Amit;

The expected number of hits is the probability of a hit (0.113) times the number of windows (2143) which is 245.6

Roger

AMIT said...

roger
If I make a statement like this

During a study of 7+ quakes of last 20 years it is found that,28% days were used as window period of the theory and 82% of 7+ quakes occured in these window period

What is wrong in the statemet?

regards
Amit

Anonymous said...

Again you still use other planets as a source of tidal influence. Using the laws of physics and assigning the suns tidal force equal to one this is the effective tidal pull of the sun, moon and planets when they are at their shortest distance from earth.
Moon 2.21
Sun 1.00
Venus 0.000113
Jupiter 0.0000131
Mars 0.0000023
Mercury 0.0000007
Saturn 0.0000005
Uranus 0.000000001
Neptune 0.000000002
Pluto 0.0000000000001
The tidal force of venus is nearly 10 times greater than Jupiter yet you still quote Jupiter as a factor.
If you added all the tidal force together (perfect planetary alignment) the tidal pull on earth would not raise the oceans tide more than a couple millimeters.
Using the moon and sun as indicators of tidal pull in earths crust is a good start but I think you are wasting everybody's time thinking Jupiter(or venus which is stronger) all on its own has an effect.
Ray

AMIT said...

hi
Philippines quake to ,6.1 ie on 20th March 2011

http://in.news.yahoo.com/quake-rocks-northern-philippines-no-damages-20110320-040952-737.html

Amit

Roger Hunter said...

Amit;

> During a study of 7+ quakes of last 20 years it is found that,28% days were used as window period of the theory and 82% of 7+ quakes occured in these window period

> What is wrong in the statemet?

Well for starters, it isn't true.

But that aside, the situation is this; suppose you were flipping a coin and reported that in 100 tosses there were 48 heads.

But how can that be? You have only one coin!

Roger

AMIT said...

Roger
think example of coin does not fit here.Without 48 coins ,we can say that
regards
Amit

AMIT said...

Roger
for that matter ,all stats of flipping the coin more than once are incorrect
Amit

Roger Hunter said...

Amit;

Mathematically, there's no difference between flipping one coin 100 times and flipping 100 coins once.

I'm trying to explain the reason why we can get more hits than we have quakes. It's not odds times quakes, it's odds times windows.

Roger

Roger Hunter said...

Amit;

Looks like the 20th supermoon is a wash.

Roger

AMIT said...

Roger
I have also used the word
28% days are used as window period-
I would like second openion about my statement -whether it is right or otherwise.
Amit

Roger Hunter said...

Amit;

I do not understand your question.

Please elaborate.

Roger

Anonymous said...

Although neither flat constant distribution nor Poisson distribution are perfect for stats work, certainly Rogers analysis is not as accurate as the Poisson/Binomial analysis I offered in previous messages and I stick to Poisson.
Remi

AMIT said...

Remi
My simple question is what is mathematically wrong in the staement given by me earlier (aa an outcome of Rogers analysis)?ie

During a study of 7+ quakes of last 20 years it is found that,28% days were used as window period of the theory and 82% of 7+ quakes occured in these window period

regards
Amit

AMIT said...

hi
Unlike Plate tectonics ,the theory does not differentiate between Earthquakes and Volcanic Eruptions.This was mentioned several times on this blog. Thus , a date for earthquakes is also prone to Eruptions

http://www.sify.com/news/600-evacuated-as-Indonesian-volcano-erupts-news-ldtrEfdebhf.html?ref=content_widget_news

When eruption occurs quake maggnitude decreases
Amit

AMIT said...

hi
A strong quake jolts japan to day

http://channel6newsonline.com/2011/03/strong-quake-strikes-off-japan-no-tsunami-warnings-issued/

amit

Roger Hunter said...

Amit;

I see the problem.

I said that you had 247 hits out of the 302 quakes which is about 82%. But that was because I was testing all your indicators individually and summing the results.

This means that one quake might be hit by several indicators. You can't compare hits to quakes, you compare hits to indicators and by that measure you got no more than expected.

Roger

Roger Hunter said...

Amit;

The main reason for the high number of hits is the high number of indicators, especially the declinations. The time from highest to lowest is only about 2 weeks so 3 day windows take up about half of the time.

The otthers fill in so about 71% of the days are covered.

Roger

AMIT said...

Roger
Expected hits 245.6 is total number of windows multiplied by odds.In that case 245.6 has nothing to do with number of inductors.
Where as 247 (total hits )depends on number of indicators.
In that case, how can we compare both? and say- Exactly as expected-
Roger I am confused.Something is missing.

Roger Hunter said...

Amit;

> Expected hits 245.6 is total number of windows multiplied by odds.In that case 245.6 has nothing to do with number of inductors.

Total number of indicators multiplied by odds on a hit.

> Where as 247 (total hits )depends on number of indicators.

Yes. The more often you try, the more hits you will get. With enough indicators you'll get them all.

> In that case, how can we compare both? and say- Exactly as expected-
Roger I am confused.Something is missing.

No, it's all there.

Roger

Anonymous said...

It does not matter if a single indicator or a whole number of indicators together. The prediction as to when there is or not an earthquake is what counts. You may use dog barkings. Who cares. Use stats to see if there is a correlation. The correlation is what MATERS. I think Amit should not tell us HOW he does it but THE RESULTING DATES. We agreed to do a test which should last for the WHOLE duration of the year. You cannot judge it the first month. Again I repeat one can use many indicators? In fact I dont care PROVIDED there is significant correlation between predicted dates AND earthquake dates.
Roger is off the mark. Sorry.
Remi

AMIT said...

Roger
That is why the process of selecting is more important ,rather than taking all indicators at a time.
With all indicators I give only 25 odd dates ie 75 odd window period only. So more the inductors ,more the hits is not correct
We should take planets changing direction as prime requirement and then any three out of rest of the indicators acting simultaneously
should be taken( ie 1 prime+ Any 3 indicators acting simultaneously-out of rest of 6)
Amit

Roger Hunter said...

Amit;

> We should take planets changing direction as prime requirement and then any three out of rest of the indicators acting simultaneously
should be taken( ie 1 prime+ Any 3 indicators acting simultaneously-out of rest of 6)

That's what is missing and was to be my next question!

Remi; How he does it is important to me so that I can duplicate his predictions and then extend them to a larger interval. Small samples can give false results.

Roger

Roger Hunter said...

Amit;

OK, when making a prediction you say day of station +-1 day if 3 other indicators are also present.

I assume you mean within the 3 day window. Otherwise same day seldom happens.

But then what are the windows 15 days either side for?

Roger

Anonymous said...

As there is no clear theory to be tested, such as Newtons Gravity, which applies to all things Predictably, here we test not so much indicators individual, as they mean nothing. An indicator on its own may not work well but in combination with other may do excellently. So we test DATES. We do not care how we got the dates. We correlate predicted dates with Quakes occured. That is the test. Trying to do individual indicators may well give below average hit rate. Together with other indicators may well allow us to predict high probability dates for quakes.
So the test should be DATES with Quake Dates.
Remi

AMIT said...

Remi
You are right,it is not pure maths as we might think.May be any one mathematical formula can not give probable dates.
All that ,we can test is dates.It could be like fine art.
Amit

AMIT said...

Roger
It may be difficult for me to explain ,exactly how I give dates.But I shall try
1) First find out period(+ or -15 days ) during which quake is possible(6+)
2)These are the dates ,when planet changes direction,not necessarily +or - 1day of stationed planet,any day between +or - 15 days
3)Once such + or - 15 days are selected ,then search for other three indicators,in those + or - 15 days.
4)From these + or - 15 days ,the dates is selected from Moon position,as Moon is the main trigger
5)Watch Moon,during these 15 days and select the date,with 3 or more inductors
6)Magnitude is decided depending on perigee of Moon,Planet distances,phases of Moon etc. More the indicators coinciding greater is the quake
Closer the planets and stronger the aspects,greater is the quake
Though this is very rough idea.There is lot of fine tuning required
Amit

Roger Hunter said...

Amit;

We have a communication problem.

I was sure you said take a 3 day window 15 days before and after a planet station.

Now you say look for 3 or more indicators in the 15 days before and after a station.

How close together must the 3 indicators be? All within a 3 day span perhaps?

Roger

Roger Hunter said...

Amit;

You say look for moon aspecting a planet station at one of the 90 degree positions.

How close? 1 degree 2 degrees, what? But don't say 30 degrees, that's unreasonable.

Roger

AMIT said...

Roger
Yes. May be I could not explain properly
1) Moon aspects are taken + or -6 degrees ie + or -12 Hrs
2)yes, Indicators are required within 3 day span. More closer they are stonger is the quake
Amit

Roger Hunter said...

Amit;

Ok, I am looking at the moon each hour for the 31 days from +15 to -15 days. If it comes within 2 degrees either way from one of the 90 degree angles (4 of them) I accept that day and mark it as a match.

Is this acceptable to you?

Since the moon moves around 14 degrees per day there are a lot of matches.

Roger

AMIT said...

Roger
Only thing is
1)2 dates from either side of stationed planet (with strong combinations ) should be selected
2)126 total stationed ie 126*4=504 widows should be selected,approximately
If we select weaker Moon combinations ,there will be lot of windows
Amit

Roger Hunter said...

Amit;

Are you saying to always select 2 dates from either side?

What about dates without other indicators?

Roger

AMIT said...

Roger
What I mean is such strong combination of indicators and aspects should be selected that yearly the dates should not exceed 30/32 odd
Yearly 6 to 8 stationed planets occur,hence 8*4=32 dates approx
Amit

AMIT said...

Roger
What I mean is such strong combination of indicators and aspects should be selected that yearly the dates should not exceed 30/32 odd
Yearly 6 to 8 stationed planets occur,hence 8*4=32 dates approx
Amit

Anonymous said...

Hi Amit,

although Roger Hunter evaluated you predictions correctly as almost random (11% vs 12-13% in total), you are right in your attempt to earthquake prediction. The gravitational pulls of Moon, Sun and planets realy trigger the earthquakes, but you mixed apples and pears together.

You must analyse the specific areas separately. Than you can see that the EQ in Chile (Feb 27, 2010) was triggered in the same tidal phase as Tokohu EQ (Mar 9+11, 2011), but with oposite directions of the pull. In the case of Chile EQ the movement of upper layers was to the west and in the case of Japan EQ the movement wes to the east.

If you will mixed both types of forces together, you obtain random results. If you will divide them according the area or focal mechanisms, you will obtain higher than random predictions. See the papers of Sachiko Tanaka when analysed only normal or reverse EQs in the subducted area.

Pavel