More results from Don's recent California earthquake forecasts:

All events in the California/Nevada area since Dec. 3 of Ml>=2.5
have been subjects of predictions by Don in Hollister. This is an
important aspect of public prediction that is often overlooked.
If a forecast system were to be adopted, the public would expect
to be warned if an event is imminent, but also would expect that
for the most part, events would not happen unless a warning had
been produced. In other words, the maximized system would have
100% hits, 0 misses and 0 false alarms. By producing forecasts
for each of the larger California events, Don is demonstrating
that his system has a fair chance of success not only in avoiding
misses and false alarms, but also in hitting most regionally significant
events.
The recent forecasts and evaluations follow:

FORECAST 1:
O: 03DEC2001 15:05:33 38.2N 122.3W Md=2.8 NCSN SW of Napa, CA
O: 03Dec2001 23:05:33 38.3N 122.3W MD=2.8 NEIS NORTHERN CALIFORNIA
P: 05-09DEC 40 km 38.4N 122.4W 2.5-3.5 Yountville, CA
http://www.earthwaves.org/wwwboard/messages/11511.html

FORECAST 2:
O: 05DEC2001 15:17:10 38.5N 119.4W Ml=2.7 NCSN ESE of Markleeville, CA
P: 27N-01DEC 40 km 38.2N 119.3W 2.8-4.1 Bridgeport, CA
http://www.earthwaves.org/wwwboard/messages/11322.html

FORECAST 3:
O: 05DEC2001 19:38:12 35.0N 119.2W ML=2.7 SCSN SE of Taft, CA
P: 03-07DEC 40 km 35.0N 118.5W 2.0-3.3 Tehachapi, CA
P: 04-08DEC 40 km 34.6N 118.9W 2.8-3.5 Santa Clarita, CA
http://www.earthwaves.org/wwwboard/messages/11353.html

FORECAST 4:
O: 04Dec2001 12:20:09 38.8N 107.4W ML=3.1 NEIS COLORADO
P: 01-05DEC 40 km 43.0N 103.0W 1.5-3.0 CHADRON, NEBRASKA
http://www.earthwaves.org/wwwboard/messages/11321.html

Notes on forecasts:

Forecast 1:
Event occurred within magnitude and distance window, but two days
later and expected. Odds of a successful prediction with a 7-day
forecast period: 1 in 17.

Forecast 2:
Event occurred within magnitude and distance window (just barely),
but 4 days later than expected. There have been 3 10-day windows in
the past 4 years which contained a "successful" event in the predicted
region and magnitude range. Odds of a successful random prediction
with a 10-day forecast period: 1 in 49.

Forecast 3:
The event occurred within the time window and magnitude range of
two forecast, however it was barely outside the distance range -
about 50 km from each forecast epicenter. In the first forecast,
16 5-day windows have seen a successful event within 50 km in the past year.
This is odds of random success of about 1 in 5 for the first forecast.
The second forecast is much less probable in a random sense because
it looks for a larger event. In the past year only 5 5-day windows have
contained a successful event within 50 km of the forecast epicenter.
Odds of random success with such a forecast are then about 1 in 73.

Forecast 4:
The forecast is clearly a miss as far as the 40 km radius is concerned,
however, the event is in the magnitude window and time window stated.
Moreover, earthquakes in this region are relatively rare, so it would
be interesting to see what the odds of success would be for an
earthquake in an area including both the forecast and the occurring
events. The number of 5-day periods occurring with Ml 1.5-3.1 in the
area 44-38N over the past year is 5. Therefore chosing a successful
5-day period containing a successful event in this area would occur randomly
about 1 time in 73 tries.

All times PST
Magnitude ranges are considered successful if event magnitude is within
0.1 units of stated magnitude range. This is the standard error
on determined Ml.