It looks like you're using an Ad Blocker.
Please white-list or disable AboveTopSecret.com in your ad-blocking tool.
Some features of ATS will be disabled while you continue to use an ad-blocker.
Originally posted by PuterMan
reply to post by mbkennel
I accept what you are saying if you are trying to forecast unusual events, however I am not attempting to do any forecasting here.
All this is observing is that the energy release levels are going up and the trendlines are showing the trend BUT I have not created any forward figures for the trend. Since this is historic data the trendline within the dataset is perfectly valid as it is not attempting to extrapolate future trends..
I agree that for forecasting one would need to go much further back, and this is an exercise that is under way, but again I doubt that it would be of any use for forecasting. The main problem being the lack of accurate older data.
I would guess you are possibly an actuary in the insurance industry?
No one has used this system successfully to forecast earthquakes for the saving of life and limb, and I would not even attempt that. Statistically EVT might be of interest to insurance companies for the purpose of assessing premiums, but really for little else.
Edit to add: The use of energy released gives a measure of the effect of plate movements since these are the major contributor to earthquake figures.
As the stress moves along a fault system so the earthquakes occur - a really good example was in Turkey where a series of quakes moved from East to West over a period of months. The total energy output in the course of a year is actually a very vague figure, but it is nonetheless an indicator of what is going on in the system as a whole. If you wanted to forecast quakes you would have to measure all the energy release in a fault system and plot the lat/lon/depth as well to get a picture of where the stresses were moving in addition to knowing the expected movement of the strike face in a given period and the friction in the system. Actually statistics have little of nothing to do with it since it is controlled by topography and plate tectonics, which is one reason why EVT will never be able to accurately predict an earthquake any better than a 'local' in the area (if it is on land).
Putting a trendline on historic data is only a visual aid to what has passed and not to the future in any way shape or form and I apologise if I gave any other impression.
In this case there is a perception that there has been an increase in the number/intensity of earthquakes. This data confirms that.
This data does NOT mean that this 'trend' will continue in 2011, or even to the end of the current year.
But when you say that the "trendline is up" there is an implicit expectation that there was in fact a positive trend.
Instead of focusing on the unstable parameter Mmax we suggest a new, stable and convenient characteristic Mmax(t) – defined as the maximum earthquake that can be recorded over a future time interval of duration t.
In fact, because EVT suggests a statistical methodology for the extrapolation of quantiles beyond the data range, the question whether such interpolation is justified or not in a given problem should be investigated carefully in each concrete situation. But EVT provides the best statistical approach possible in such a situation
Statistically EVT might be of interest to insurance companies for the purpose of assessing premiums, but really for little else.
a geophysicist who had deep earthquake and data set knowledge
Evidence of a statistically significant trend might encourage one to look and search for potential physics behind it.
Originally posted by bluedrake
Just one quick question though, surely a release of energy in smaller 7-8 quakes is a good thing. I think if there was almost no energy been release it would mean that its storing up for the big one...
Looks like the big one was in 2004.... maybe we will get another big one before the energy release starts to decline