[quote=ewmayer;117554]IIRC decibel [as applied to acoustics] multiplies the sound energy [actually the power-per-sq-meter, I believe, since the total energy and power depend on the areal extent of the acoustic front] by a factor of 10, so that scale also has an arbitrary constant factor mixed in.[/quote]
The factor of 10 here isn't "arbitrary". It follows from the bog standard
definition of the prefix "deci".

[QUOTE=davieddy;117562]The factor of 10 here isn't "arbitrary". It follows from the bog standard definition of the prefix "deci".[/QUOTE]

The choice of *scaling* is arbitrary - but as you note, once one has chosen a scaling, the resulting unit *naming* is not. I was referring to the former, not the latter.

I still think log base 32 is a load of bollocks.

[QUOTE=davieddy;117571]I still think log base 32 is a load of bollocks.[/QUOTE]

Decimal fascist. :mad:

[quote=ATH;117532]The richter scale is ~ log[sub]10[/sub](amplitude) but the energy rises a factor of 1000 when richter scale rises 2 values, so the Richter scale is ~ log[sub]sqrt(1000)[/sub](energy) ~ log[sub]32[/sub](energy).

[/quote]

Is "Richter" case sensitive here?
I guess now that the factor of 1000 is an empirical observation.
I am having difficulty squaring this with intensity(W/m^2) being
proportional to amplitude^2.

David

[quote=ewmayer;117554]
In one way, expressing earthquake magnitude as log10(amplitude) actually makes prefect sense, since the shaking you feel is proportional to wave amplitude, not total-energy-release. But again, it all comes down to picking one's arbitrary logarithmic multiplier.[/quote]

One big thing in favour of using intensity(W/m^2) instead of
amplitude is conservation of energy.

[quote]
Originally Posted by [B]ewmayer[/B] [URL="http://www.mersenneforum.org/showthread.php?p=117554#post117554"][IMG]http://www.mersenneforum.org/images/buttons/viewpost.gif[/IMG][/URL]
[I]IIRC decibel [as applied to acoustics] multiplies the sound energy [actually the power-per-sq-meter, I believe, since the total energy and power depend on the areal extent of the acoustic front] by a factor of 10, so that scale also has an arbitrary constant factor mixed in.[/I]
[/quote][quote=davieddy;117562]The factor of 10 here isn't "arbitrary". It follows from the bog standard
definition of the prefix "deci".[/quote]
deci=10^-1, deca (or deka) = 10^1
IIRC, bels were inconveniently large in some situations, so dBs were used. Makes sense considering it was originally used ([I]inter alia[/I]) to measure S/N on standard telephone lines. Also the use of dB implies a ratio, rather than an absolute value (of course, sometimes the baseline reference value is 0).

No, the baseline reference can't be 0... sheesh. Too much typing, not enough thinking.

[QUOTE=davieddy;117652]One big thing in favour of using intensity(W/m^2) instead of amplitude is conservation of energy.[/QUOTE]

Depends on one's POV - personally, during an earthquake, I'm more concerned about whether my house is going to fall down on top of me than whether total seismic energy is being conserved - but I'm probably just weird that way.

OTOH geologists obviously do care very much about total energy release and where that energy winds up going.

The folks at the USGS did say that this latest quake along the normally-quiete Calaveras fault increases the chance of a major quake along the nearby Hayward fault, due to strain displacement - but only by a small percentage, relative to the previously estimated ~25% odds of a major quake along the Hayward in the next 25 years:

[url]http://seismo.berkeley.edu/hayward/probabilities.html[/url]

The above answers michaf's earlier question: yes, it seems likely we're gonna have another big one in the next few decades:
[quote]A new probability report was released in 2003 by Working Group '02. This report concludes that the Bay Area faces a 62% probability of an M6.7 or larger earthquake over the next 30 years (2003-2032).[/quote]

===

[i]Aside:
While I don't question the overall probability, I am skeptical of the claimed uncertainty of 10%, which seems ludicrously low to me. Obviously it's possible that no such quake occurs in the given time frame, which instantly gives a much-larger -62% negative uncertainty, and similarly if such a quake does occur, we get a +38% uncertainty. Maybe I'm misunderstanding what they mean by "+-10%", it's probably something like the one-sigma confidence interval of their computer models - but as with weather prediction, that ignores "the uncertainty in the uncertainty," i.e. there are many assumptions built in such complex-systems models, if one proves significantly wrong, all your earlier uncertainty estimates are right out the window.[/i]

[quote=ewmayer;117471]
Anyway, admit it: you're just jealous of my pulling-numbers-out-of-my-butt quake postdiction skills. ;)[/quote]

Could you address these skills to the "fact" that a quake
with 32 times greater energy release produces waves with
a peak amplitude 10 times greater (power 100 times greater)?

The word "coupling" comes to mind.

PS Brainwave(?): isn't power freqency dependent as well?
If so perhaps bigger quakes have lower frequency.
Yes intensity proportional to density*wavespeed*(amplitude*frequency)^2

[QUOTE=ewmayer;117664]
[i]Aside:
While I don't question the overall probability, I am skeptical of the claimed uncertainty of 10%, which seems ludicrously low to me. Obviously it's possible that no such quake occurs in the given time frame, which instantly gives a much-larger -62% negative uncertainty, and similarly if such a quake does occur, we get a +38% uncertainty. Maybe I'm misunderstanding what they mean by "+-10%", it's probably something like the one-sigma confidence interval of their computer models - but as with weather prediction, that ignores "the uncertainty in the uncertainty," i.e. there are many assumptions built in such complex-systems models, if one proves significantly wrong, all your earlier uncertainty estimates are right out the window.[/i][/QUOTE]

If the quake doesn't fall into the 30 year time-frame, the probability could of course still have been 62%. The time-frame just happend to be entirely in the other 38%.

But still, it's a very major chance anyway you look at it. I'm glad to live in Holland, albeit much of that is below sealevel. (Flooding is the only major disaster that can strike Holland afaik)