Forecast accuracy
Posted by EQF on May 16, 2009 at 05:18:26:

Hi Boyko,

The type of forecast information you are generating is highly compatible with my own forecasting program. And I would be interested in adding your data to my program so that it would display if the appropriate option were chosen. However, before going through that trouble of adding those data I would need to know how well your forecasts have matched past earthquakes. Could you do the following? This should not take you too long.

Get a list of the 20 or 30 most powerful earthquakes that occurred during the past year. They are available with searches at a USGS Web site. Organize them by magnitude with the highest magnitude earthquake at the top.

Below each earthquake in the list, place a copy of the forecast data from your prediction lists that you think best matched the earthquake. Don’t worry about the magnitude of the forecast. If you believe that one of several forecast lines might have matched the earthquake and are not sure which one it was, list them all.

Post the results to this bulletin board and I will look at them and probably post an opinion regarding how helpful I think the data might be.

My own forecasting program is scheduled to add routines that do what it appears to me your own routines might be doing. And that is to calculate higher probability time, latitude, and longitude windows based on Solid Earth Tide and Tide Generating Force information. The equations for doing that need to be created. If the equations that you are using already work for that then perhaps they could be incorporated in the program.

Thanks.


Follow Ups:
     ● Re: Forecast accuracy - Skywise  20:56:12 - 5/16/2009  (75250)  (1)
        ● Re: Forecast accuracy - EQF  07:04:42 - 5/18/2009  (75297)  (1)
           ● Re: Forecast accuracy - Skywise  22:02:01 - 5/18/2009  (75305)  (0)
     ● Re: Forecast accuracy - Roger Hunter  13:46:31 - 5/16/2009  (75249)  (1)
        ● Re: Forecast accuracy - EQF  06:39:57 - 5/18/2009  (75296)  (0)