+- Earthwaves Earth Sciences Forum (http://www.earthwaves.org/forum)
+-- Forum: Earthwaves (http://www.earthwaves.org/forum/forumdisplay.php?fid=16)
+--- Forum: Earthquake Predictions (http://www.earthwaves.org/forum/forumdisplay.php?fid=8)
+--- Thread: Nabhan final results (/showthread.php?tid=128)
Nabhan final results - Roger Hunter - 08-10-2014
I've finally finished evaluating David Nabhan's prediction ideas, thanks to considerable help from Skywise and helpful comments from Dr Vidale.
Nabhan's claims of correlations with moon phases and perigees are simply wrong. The hits on those things are totally at chance level.
Dawn and dusk hits were another matter. They happen far more often than chance would allow and when that happens, something is amiss.
Aftershocks were the obvious choice. They don't follow normal rules but the usual method of elimination didn't solve the problem. There were not that many big quakes in the area.
What finally worked was to eliminate quakes within a day after any quake that was aso within a degree distant from the previous quake.
What was happening was small sequences of mag 5 quakes at random times lasting long enough to hit one of the windows. Once these were removes, the peaks vanish.
If anyone is interested in seeing the actual data that supports my conclusion, please contact me. It's a bit too much to post here.
Roger
RE: Nabhan final results - Island Chris - 08-10-2014
Hi Roger,
I don't want the full results, but your post does not make clear why M5+ quakes that you remove would be at dawn or dusk. I'd say that if there are enough of these quakes and they show a significant pattern, then that means something: either your analysis is flawed or there is some chance that something seismologic is going on.
Chris
RE: Nabhan final results - Roger Hunter - 08-10-2014
(08-10-2014, 11:41 AM)Island Chris Wrote: Hi Roger,
I don't want the full results, but your post does not make clear why M5+ quakes that you remove would be at dawn or dusk. I'd say that if there are enough of these quakes and they show a significant pattern, then that means something: either your analysis is flawed or there is some chance that something seismologic is going on.
Chris
Hi Chris
They aren't necessarily at dawn or dusk but if there's a series, some of them will be in there. Once a peak gets started it will grow as more quakes are added. Also, if you start the search at 2000 there's no peak at all.
Roger
(08-10-2014, 02:15 PM)Roger Hunter Wrote:(08-10-2014, 11:41 AM)Island Chris Wrote: Hi Roger,
I don't want the full results, but your post does not make clear why M5+ quakes that you remove would be at dawn or dusk. I'd say that if there are enough of these quakes and they show a significant pattern, then that means something: either your analysis is flawed or there is some chance that something seismologic is going on.
Chris
Hi Chris
They aren't necessarily at dawn or dusk but if there's a series, some of them will be in there. Once a peak gets started it will grow as more quakes are added. Also, if you start the search at 2000 there's no peak at all.
Roger
I should add that the deletions are not specific to the windows, The deletion starts when a quake is the first of a series, regardless of the time.
Roger
RE: Nabhan final results - Roger Hunter - 08-11-2014
I should know better than to say "Final".
Found another bug.
More later.
Roger
RE: Nabhan final results - Roger Hunter - 08-12-2014
(08-11-2014, 06:10 PM)Roger Hunter Wrote: I should know better than to say "Final".
Found another bug.
More later.
Roger
Ok, back on track.
Perigee and Moon phase are not significantly above or below chance.
The dawn window is far beyond chance if all files in the area are included and PST time is used.
However, this can be changed by selecting different boundaries for the area, by using true local time for the quakes or by eliminating quakes that are close together in space and time.
This all goes to show that this is not a general rule which would apply to any subset of the quakes but rather an anomalous result depending on which quakes or time standards are chosen.
Roger
RE: Nabhan final results - Roger Hunter - 08-15-2014
(08-12-2014, 03:13 AM)Roger Hunter Wrote:(08-11-2014, 06:10 PM)Roger Hunter Wrote: I should know better than to say "Final".
Found another bug.
More later.
Roger
Ok, back on track.
Perigee and Moon phase are not significantly above or below chance.
The dawn window is far beyond chance if all files in the area are included and PST time is used.
However, this can be changed by selecting different boundaries for the area, by using true local time for the quakes or by eliminating quakes that are close together in space and time.
This all goes to show that this is not a general rule which would apply to any subset of the quakes but rather an anomalous result depending on which quakes or time standards are chosen.
Roger
If anyone wants the final report, it's attached here as a .pdf file.
Comments welcome.
Roger
RE: Nabhan final results - Roger Hunter - 08-16-2014
(08-15-2014, 07:13 PM)Roger Hunter Wrote:(08-12-2014, 03:13 AM)Roger Hunter Wrote:(08-11-2014, 06:10 PM)Roger Hunter Wrote: I should know better than to say "Final".
Found another bug.
More later.
Roger
Ok, back on track.
Perigee and Moon phase are not significantly above or below chance.
The dawn window is far beyond chance if all files in the area are included and PST time is used.
However, this can be changed by selecting different boundaries for the area, by using true local time for the quakes or by eliminating quakes that are close together in space and time.
This all goes to show that this is not a general rule which would apply to any subset of the quakes but rather an anomalous result depending on which quakes or time standards are chosen.
Roger
If anyone wants the final report, it's attached here as a .pdf file.
Comments welcome.
Roger
A final note;
There's a list of dates to watch in his book covering 2011 to 2020.
I checked the period from 2011 to mid-July 2014 which included 53 predicted dates and he got a grand total of 10 hits for mag 5+ quakes within 3 days of his predicted date.
Sounds about right - for chance.
Roger