I-MAG STS    Corporation
      
      In the three months prior to the 8.7 on March 
28 there was an average of one 4.45 
earthquake every 5 days. Then there were 120  
earthquakes on the March 28th alone (purple 
bar). To be sure, more than 80 were less than 
5.0 and only there was only one 6.0. The next 
day there were 19 aftershocks in the low 5s 
and almost 150 in the 4s. On the 30th 5 
aftershocks in the 5s and 64 in the middle 4s.  
If one copies a column of a matrix and runs a 
correlation calculation the answer will 
trivially be 1. However, sometimes by shifting 
the copied data up one or more rows (lagging) 
a serial correlation (also known as 
autocorrelation) can be found such that a 
variable's past values can predict future    
      
      values. Unfortunately, the correlation with past earthquakes is miserable. Ironically, were one to run time backwards the 
aftershocks form a fine indicator that a large earthquake happened - at least on Nias. Our preliminary results indicate that 
the next major earthquake will be in the straits between Sumatra and the Mentawai Islands. There are 70 islands in the 
group so we will refine the location to near Siberut, Sipou amd North Pagai. These islands are roughly on a northwest 
southeast diagonal between 1 and 3 degrees south and 98.5 and 100.5 degrees east.  We are currently trying to assess the 
significance of the three recent large earthquakes:  6.7 on September 20 at 8:31 (UTC); 7.0 on September 13 at 3:35 and 7.9 on 
September 12 at 23:49.   More to follow including semi-divisive non-Euclidean cluster analysis soon.