Di hvdrogr Z 40, 198" H.3 Klein, Benthic «tonus 
99 
Figure 8 shows the result of a simulation running from October 6 until October 25, i. c. 
the time the cyclonic meddy passes the polygon. At 200 m a. b. all particles are trapped by 
the meddy. At 10 m a ta. more than half of the particles released arc still inside the nicdtly 
The cyclonic sense of rotation is c\iclcnt. 70 m a. b. the particles leave the polygon in a 
westward direction. This may be due either to the curvature of the meddy axis or to the shift 
of the core position at the pycnoclinc between BBL and the interior ol the deep sea. 
The simulation in Fig 9 begins on December 10 During this time no eddies arc 
observed in the area; the circulation is representative for the mean values about the whole 
experiment. In each level the particles leave the poly gon in a westward direction. Approa 
ching towards the bottom, the flow rotates cydomcally 
Deformation clue to shear 
Le Croupe Tourhillon demonstrated the deformation of a patch of Mediterranean 
Water due to shear in the Tourhillon Eddy The patch was carried around the eddy 
anticyclonicallv and was drawn out into thin filaments. Finally it w as split into two parts. To 
demonstrate how this kind of deformation works near the bottom of the deep sea. the 
deformation of a small grid due to the quasi-lagrangian transport of the points establishing 
the grid is shown in Fig. 10. The numbers near the grids give the simulation time in days. 
Again, the simulation begins on October 6 (transit of the meddy). The cyclonic rotation of 
the grid and its stretching after a few days is obvious. The area of the grid must remain 
constant, because streamfunction ipis non-divergcnt. The points which leave the polygon 
are lost, so the simulation can only last a couple of days. In all likelihood the grids would he 
stretched into very thin filaments after only a few rotations. This is a condition necessary for 
small-scale mixing processes to work effectively and to smooth out the gradients of any 
inhomogencotisly distributed matter or hydrographic properly between the patch and the 
surrounding water mass; fore.xample salinity or the concentration of nutrients 
Conclusion 
The benthic storms between September '83 and September '84 obviously occur in 
connection with deep reaching synoptic eddies fn all probability most of the other storms in 
Table 1 appear also due to deep reaching vortices. Naturally not every vortex causes a 
benthic storm according to our definition, even though a slight increase of the near bottom 
velocity can always be observed. The eddies transmit a part of their kinetic energy to the 
BBL and the resulting increase of bottom friction enhances the height of the BBL. Such 
events are essential for particles which arc released at the seabed or within the BBL. The 
pulses of high velocity flow enable the particles to cross the interface between BBL and the 
dcepsca. R. obi ns on and Kupfcrman (1985| suggested 3 mechanisms for this transfer: 
1) The dismantling of the pycnoclinc at the lop of the BBL, i. e. the reduction of the 
gradients of the hydrographic properties between BBL and deep sea. 
2) The dctatchmcnt of parts of the BBL; or. 
3) The breaking of international waves at the pycnoclinc 
Fig. II shows 3 CTD profiles with potential temperature 0 and particle concentra 
tion C. The probes arc taken at nearly the same position in the central NO AMP area, but at 
different times, i e September 'S3. ’84. and 85. The profiles A and B dearly show the 
sharp interface between the well-mixed BBL and the stratified interior of the deep sea. In 
profile C this interface is eroded. The profiles were taken by F. Nyffcler and Ch.-H Godct, 
University Ncuchatel. who participated in the NOAMP cruises (Nyffcler and Godct 
11986]) 
Another transfer mechanism is given direct by the vortex. Matter captured by the eddv 
is forced to follow its movement Svnoptic eddies are known to move for several months 
without any significant exchange belvveencore water and the surrounding water mass.