Ayecarumba Joined: Nov 17, 2009
• Posts: 6763
July 16th, 2019 at 4:02:11 PM permalink
In the days following the recent magnitude 7.1 earthquake in California, the U.S. Geological Survey put out the following statement:

Quote: U.S.G.S.

"...the chance of an earthquake of magnitude 7 or greater from the July 5 earthquake is 1 in 300..."

Just what does this mean?

"If the conditions that caused the 7.1 were replicated 300 times, they would be followed by a magnitude 7 or greater quake once."

or

"Of all the past events where a 7.1 earthquakes were recorded, one out of 300 was followed by a magnitude 7 or greater quake."

Apparently, there is a model that they are working from, but just how do they develop the odds? By not giving a timeframe, are they hedging? Maybe the event could happen 100 years from now?
Simplicity is the ultimate sophistication - Leonardo da Vinci
Boz Joined: Sep 22, 2011
• Posts: 5701
Thanks for this post from: July 16th, 2019 at 4:08:25 PM permalink
Quote: Ayecarumba

In the days following the recent magnitude 7.1 earthquake in California, the U.S. Geological Survey put out the following statement:

Quote: U.S.G.S.

"...the chance of an earthquake of magnitude 7 or greater from the July 5 earthquake is 1 in 300..."

Just what does this mean?

"If the conditions that caused the 7.1 were replicated 300 times, they would be followed by a magnitude 7 or greater quake once."

or

"Of all the past events where a 7.1 earthquakes were recorded, one out of 300 was followed by a magnitude 7 or greater quake."

Apparently, there is a model that they are working from, but just how do they develop the odds? By not giving a timeframe, are they hedging? Maybe the event could happen 100 years from now?

Need context on your quote to a time frame like you said. Of course another 7+ will happen...eventually. We can only bet on lessor ones in a shorter timeframe.
Wizard Joined: Oct 14, 2009
• Posts: 21778
Thanks for this post from:  July 16th, 2019 at 4:20:44 PM permalink
There must have been some time and geographic constraints on that quote. Since earthquakes can to some degree become overdue, I think discussing the odds of them is complicated.
It's not whether you win or lose; it's whether or not you had a good bet.
Ayecarumba Joined: Nov 17, 2009
• Posts: 6763
July 16th, 2019 at 5:11:41 PM permalink

Quote: USGS FAQ

Are earthquake probabilities or forecasts the same as prediction?
No. Probabilities and forecasts are rather like climate probabilities and weather forecasts, while predictions are more like statements of when, where, and how large, which is not yet possible for earthquakes.

Probabilities describe the long-term chances that an earthquake of a certain magnitude will happen during a time window. Most earthquake probabilities are determined from the average rate of historical events. Assuming the annual rate is constant, one can make a probability statement about the likelihood of such an event in the next so-many years. These probabilities might range from 1-in-30 to 1-in-300.

For some faults, historical occurrences are not available, but rate of slip along the fault can be estimated. Assuming a particular magnitude, one can estimate the number of years it would take to accumulate the required amount of slip. This estimate can be used to give an annual rate and used in the same manner as historical rates. These probabilities might range from 1-in-300 to 1�in-3000.

Forecasts are like probabilities but for shorter time windows, and we generally apply this term to aftershocks. After a large earthquake, there are aftershocks that are typically less frequent and smaller over time. Most aftershock sequences follow the same pattern, so the probability of an aftershock in a time window following an earthquake can be determined. These probabilities might be larger than 1-in-30.

So the article I quoted in the OP left out the important "time window" detail. Based on this FAQ it appears that they are using past performance to predict future events.
Simplicity is the ultimate sophistication - Leonardo da Vinci
unJon Joined: Jul 1, 2018
• Posts: 1738
July 16th, 2019 at 5:29:46 PM permalink
Quote: ayecarumba

Based on this FAQ it appears that they are using past performance to predict future events.

At some level, that�s always how the future is predicted.
The race is not always to the swift, nor the battle to the strong; but that is the way to bet.
Ace2 Joined: Oct 2, 2017
• Posts: 598
July 17th, 2019 at 1:59:50 AM permalink
Check out the Gutenberg-Richter law.

If I understand it correctly, it would state that if a 4.6 earthquake occurs, on average, once per period in that region, then the probability of a 7.1+ quake during the same period would be about 1 in 316.

316 being 10^(7.1 - 4.6) and assuming a �b� value of 1. In which case 316 is the same as the magnitude ratio.
Last edited by: Ace2 on Jul 17, 2019
It�s all about making that GTA
Wizard Joined: Oct 14, 2009
• Posts: 21778
July 17th, 2019 at 3:44:50 AM permalink
Quote: Ace2

Check out the Gutenberg-Richter law.

If I understand it correctly, it would state that if a 4.6 earthquake occurs, on average, once per period in that region, then the probability of a 7.1+ quake during the same period would be about 1 in 316.

316 being 10^(7.1 - 4.6) and assuming a �b� value of 1. In which case 316 is the same as the magnitude ratio.

That's quite interesting. It would suggest if you multiply the power of an earthquake by its probability, for given interval of time, you get a constant.

However, since they are due to moving plates, I assume that short-term probability will depend on recent history to some degree.
It's not whether you win or lose; it's whether or not you had a good bet.
billryan Joined: Nov 2, 2009