A second slower rate is more arbitrarily defined by the data at the end 
of the monitoring program. The linear regression of cruises 8-12 gives a half 
life of 3.4 years (r=-.40) as shown in figure 20, but cruises 9-12 did not 
predict a decrease with time. The average inventory of barite between 0.5- 
2 km from the drill site during cruises 9-12 is 5.540.1 (x10*) lb. This value 
is 12 percent higher than the average predrilling inventory of 4.940.1 (x104) 
1b calculated for cruises 1 and 2 (fig. 19). The total inventory (fig. 19) 
should be monitored again at annual intervals to determine the slower rate of 
removal. 
A similar calculation of the change in net barite inventory of the fine 
fraction (fig. 21) yields a rapid initial half-time of 0.25 years. The 
secondary rate after cruise 8 appears much slower, or zero; there are 
insufficient data for an estimate. 
This model is certainly oversimplified, but retains some merit. One 
basic assumption of this exponential decay model is that each particle has the 
same probability of escaping. In this case, the mechanism of escape is 
thought to be suspended or bed-load transport away from the drill rig or 
downward mixing and exchange with uncontaminated sediments below. Although 
larger barite particles generally would have a lower probability of eroding 
than finer particles, there are relatively few barite particles large enough 
to resist sediment transport by resuspension. Bottom stresses on Georges Bank 
are frequently greater than 1.7 dynes/cm2, the stress required to resuspend 
barite particles 63 um in diameter (Butman and Moody, 1983). Since about 
96 percent of the barite used in drilling is finer than this size, we assume 
that most of the whole size range of barite particles will be moved frequently 
by resuspension. Therefore, we attribute the initial rapid rate of Ba removal 
to the effects of resuspension and transport within the water column. 
63 
