112
L. G.-Weil et al.
An interesting aspect of the observations above is the apparent change in shape
of the long-lived eddies A1 and A2, when they moved away from their generation
area. Even if anticyclones are generally recognized as high-termperature (and lowchlorophyll) eddies, on occasion the converse also appears to be true. Accordingly,
both A1 and A2 seem to exhibit a cold core in Fig. 5.7a, b. This is due to the fact
that at this time the high insolation and the low winds render offshore surface waters
warmer than that transported by the eddies. Indeed, the same eddies displayed again
a high-temperature (and low-chlorophyll) core 1 month later (Figs. 5.7c and 5.8a–e),
when convection produced by surface cooling, aided by wind stirring, weakened the
seasonal stratification. This resulted in colder offshore waters, with higher phytoplankton content than that observed in A1 and A2. These changing environmental
conditions could be also related to modifications of the sub-mesoscale regime in
which the mesoscale features, A1 and A2, are embedded. The Chl-a images of January and February 1999 (Fig. 5.8a, b) show a sub-mesoscale field dominated by short
wavelength features that complicate even the mere identification of A1 and A2 in the
satellite images. The spatial scales of this sub-mesoscale field increased gradually
with time, as shown in the SeaWiFS scenes of April 1999 (Fig. 5.8d, e), in which A1
and A2 are easily observed.
A third anticyclonic eddy, labeled A3, was observed southwest of La Palma island
on September 16, 1998. Opposite to the previously described eddies, which separate
from their generation area in approximately 2 weeks, A3 was retained in the same
position close to La Palma island until the end of November (i.e. more than 70 days),
displaying a constant diameter of ∼ 80 km (Figs. 5.6c and 5.7a). Afterwards, it shifted
quickly to the west (Fig. 5.7b), at an estimated mean speed of ∼ 6.5 cm s
− 1 (17 day
average), slowing down subsequently, and being observed last on December 26,
1998. This anticyclonic eddy was never coupled to a counter-rotating cyclonic eddy,
for the whole period in which it was attached to La Palma island (unlike what occurs
in the leeway of an isolated obstacle when leeside attached eddies appear). Other
works reporting the presence of trapped eddies, downstream of isolated islands and
seamounts, argue that sometimes incident flows are not strong enough to generate a
vortex street of eddies (Bograd et al. 1997). A similar explanation could be given for
the prolonged (September to November) attachment of A3 southwest of La Palma.
Another long-lived cyclonic eddy, labeled C3 and located over the bathymetric
trough downstream of the shallower channel between Fuerteventura and Africa, was
observed for the first time on June 27, 1998 (Fig. 5.1a). Unlike the other eddies
described earlier, C3 was not generated and shed from one of the islands. Its origin
is thought to be related to potential vorticity conservation, as the alongshore flow
encounters the deeper water of the trough, producing such an eddy as long as the flow
through the zone is equatorward (Barton et al. 1998, 2004). C3, detected in Chl-a, BT
and SLA observations, showed a mean diameter of ∼ 90 km and initially remained
stationary around 27.2
◦ N and 15
◦ W (Figs. 5.3d–h, 5.4c–e and 5.6a). In September,
it started to shift westward, approximately at the same time as A1 began to move
offshore (Figs. 5.4f and 5.6b, c). One month later, C3 was located south of Gran
Canaria (Figs. 5.4g, h and 5.7a). The last evidence of C3 in passive radiometer data
occurred on November 19. It was still observed, moving southwestward (Fig. 5.4i),
in SLA maps until February 10, 1998.
L. G.-Weil et al.
An interesting aspect of the observations above is the apparent change in shape
of the long-lived eddies A1 and A2, when they moved away from their generation
area. Even if anticyclones are generally recognized as high-termperature (and lowchlorophyll) eddies, on occasion the converse also appears to be true. Accordingly,
both A1 and A2 seem to exhibit a cold core in Fig. 5.7a, b. This is due to the fact
that at this time the high insolation and the low winds render offshore surface waters
warmer than that transported by the eddies. Indeed, the same eddies displayed again
a high-temperature (and low-chlorophyll) core 1 month later (Figs. 5.7c and 5.8a–e),
when convection produced by surface cooling, aided by wind stirring, weakened the
seasonal stratification. This resulted in colder offshore waters, with higher phytoplankton content than that observed in A1 and A2. These changing environmental
conditions could be also related to modifications of the sub-mesoscale regime in
which the mesoscale features, A1 and A2, are embedded. The Chl-a images of January and February 1999 (Fig. 5.8a, b) show a sub-mesoscale field dominated by short
wavelength features that complicate even the mere identification of A1 and A2 in the
satellite images. The spatial scales of this sub-mesoscale field increased gradually
with time, as shown in the SeaWiFS scenes of April 1999 (Fig. 5.8d, e), in which A1
and A2 are easily observed.
A third anticyclonic eddy, labeled A3, was observed southwest of La Palma island
on September 16, 1998. Opposite to the previously described eddies, which separate
from their generation area in approximately 2 weeks, A3 was retained in the same
position close to La Palma island until the end of November (i.e. more than 70 days),
displaying a constant diameter of ∼ 80 km (Figs. 5.6c and 5.7a). Afterwards, it shifted
quickly to the west (Fig. 5.7b), at an estimated mean speed of ∼ 6.5 cm s
− 1 (17 day
average), slowing down subsequently, and being observed last on December 26,
1998. This anticyclonic eddy was never coupled to a counter-rotating cyclonic eddy,
for the whole period in which it was attached to La Palma island (unlike what occurs
in the leeway of an isolated obstacle when leeside attached eddies appear). Other
works reporting the presence of trapped eddies, downstream of isolated islands and
seamounts, argue that sometimes incident flows are not strong enough to generate a
vortex street of eddies (Bograd et al. 1997). A similar explanation could be given for
the prolonged (September to November) attachment of A3 southwest of La Palma.
Another long-lived cyclonic eddy, labeled C3 and located over the bathymetric
trough downstream of the shallower channel between Fuerteventura and Africa, was
observed for the first time on June 27, 1998 (Fig. 5.1a). Unlike the other eddies
described earlier, C3 was not generated and shed from one of the islands. Its origin
is thought to be related to potential vorticity conservation, as the alongshore flow
encounters the deeper water of the trough, producing such an eddy as long as the flow
through the zone is equatorward (Barton et al. 1998, 2004). C3, detected in Chl-a, BT
and SLA observations, showed a mean diameter of ∼ 90 km and initially remained
stationary around 27.2
◦ N and 15
◦ W (Figs. 5.3d–h, 5.4c–e and 5.6a). In September,
it started to shift westward, approximately at the same time as A1 began to move
offshore (Figs. 5.4f and 5.6b, c). One month later, C3 was located south of Gran
Canaria (Figs. 5.4g, h and 5.7a). The last evidence of C3 in passive radiometer data
occurred on November 19. It was still observed, moving southwestward (Fig. 5.4i),
in SLA maps until February 10, 1998.