Earthquake Pakistan 8 Oct 2005
A devastating earthquake hit Pakistan in the mountainous regions including the Kashmiri border area. Estimates of the death toll range from 20,000 to 30,000, but remain uncertain. The area is very difficult to access, so rescue efforts are still slow at the time of this writing three days later. The main temblor was on Saturday, October 08, 2005 at 03:50:38 UTC, and it was a huge quake, Richter magnitude 7.6.
The preliminary release by the United States Geological Survey, National Earthquake Information Center on the 8th gives basic information: "A major earthquake occurred about 95 km (60 miles) north-northeast of Islamabad, Pakistan at 9:50 PM MDT, Oct 7, 2005 (Oct 08 at 8:50 AM local time in Pakistan). The magnitude and location may be revised when additional data and further analysis results are available. No reports of damage or casualties have been received at this time; however, this earthquake may have caused substantial damage and casualties due to its location and size. The magnitude was furnished by the USGS National Earthquake Information Center."
The GCP formal prediction was for a period beginning 30 minutes before the main temblor and continuing for 8 hours. The full network result was significant at Z=1.773 (.95*1.866), and p=0.038. The first figure below shows this result.
The second figure is the same Stouffer Z based analysis for the full GMT day to provide some context. It shows a striking shape (that likely is just chance fluctuation, of course) around the time of the quake. But there is a strong trend with significant slope beginning at the time of the quake that continues for a few hours beyond the formal analysis period.
Because we have seen characteristic patterns in the variance measure for some large and destructive quakes, this possibility is explored in the next figure. It shows the cumulative deviation of the variance from expectation for an 8 hour period centered on the Pakistan quake. As was the case for the huge quake that caused the Tsunami in Dec 2004, this picture shows a downward spike with a minimum near the time of the main temblor. Related analyses indicate that the GCP network tends to show a characteristic response to big quakes that occur where human populations are affected.