428
thermocline circulation on coastal upwelling.
An important conclusion is the existence of a narrow and intense baroclinic coastal jet,
associated to the upwelling region, which is indeed the boundary condition for the interior
CC. This jet reaches its maximum where the seasonal thermocline approaches the surface
but is embedded over a larger region due to the raise of the permanent thermocline,
which prevents it from being a spatially and temporally localized current. Furthermore,
as pointed out in the previous subsection, the inflow from the CC into the coastal region
must recirculate southward near the African coast. Such an inflow from the CC would
guarantee the continuity of the coastal jet even during periods of no local upwelling. Fig. 6
shows the mean and first temporal Empirical Orthogonal Functions (EOF) resulting from
the analysis of eight years of SST data. The mean EOF clearly shows a latitudinal SST
gradient but it is also suggestive of flow continuity between the CC and the coastal region.
The first EOF shows maximum seasonal variability both in the north (seasonal warming)
and the coastal region between Cape Vert and Cape Blanc (seasonal upwelling).
The band of heat flux density in the coastal upwelling region is visible in Fig. 3 (see also
Schmitt et al. 1989). This band is much wider than the coastal upwelling region, partially
due to the existence of an extense upward Ekman pumping region (which reaches 200-300
km from the coast in areas adjacent to capes: Bakun and Nelson 1991, McClain and
Firestone 1993) and to the existence of a wide region of enhanced mixing associated to
the coastal jet (Foo 1981, Kundu and Beardsley 1991, Pelegrf and Richman 1993). The
main reason, however, is probably related to the upwelled water moving into warmer
interior regions where it continues gaining heat. We may estimate the heat flux density
necessary to warm the upwelling water as pCp(Oo - Oi)uehl L, where ueh is the offshore
Ekman transport per meter of coastline, L is the width of the upwelling region, and Oi and
00 are the surface temperatures of the upwelling and offshore areas, respectively. Using
(}i = 18°C, (}o = 23°C, L = 250 km and ueh = 1 m 2 Is gives 80 W 1m 2 • Fig. 1 shows the
remarkable heat flux density from the atmosphere in the coastal upwelling region at 23°N,
the highest in the North Atlantic, as obtained from weather data; the annually averaged
heat influx density, as calculated from Fig. 1, is Qn = 140 W 1m 2 • Comparing these two
values we find that heat transport divergence in the coastal upwelling region is rather large,
about 60% the heat influx density. The heat transported away from the upwelling region
due to offshore Ekman transport can be estimated multiplying our previous estimate by
L which gives 1.1 x 10 13 W, about 25% the annually averaged heat influx in the AC-CC
region. During winter, however, heat influx associated to coastal upwelling is the only
mechanism that continues operating. Fig. 1 also shows, for comparison, the heat flux
densities in the trade region, slightly north of the intertropical convergence zone (ITZC);
in this case the incoming radiation flux density is also high, although not as much as in
the coastal upwelling region, but loss through latent heat flux density is comparably large.
thermocline circulation on coastal upwelling.
An important conclusion is the existence of a narrow and intense baroclinic coastal jet,
associated to the upwelling region, which is indeed the boundary condition for the interior
CC. This jet reaches its maximum where the seasonal thermocline approaches the surface
but is embedded over a larger region due to the raise of the permanent thermocline,
which prevents it from being a spatially and temporally localized current. Furthermore,
as pointed out in the previous subsection, the inflow from the CC into the coastal region
must recirculate southward near the African coast. Such an inflow from the CC would
guarantee the continuity of the coastal jet even during periods of no local upwelling. Fig. 6
shows the mean and first temporal Empirical Orthogonal Functions (EOF) resulting from
the analysis of eight years of SST data. The mean EOF clearly shows a latitudinal SST
gradient but it is also suggestive of flow continuity between the CC and the coastal region.
The first EOF shows maximum seasonal variability both in the north (seasonal warming)
and the coastal region between Cape Vert and Cape Blanc (seasonal upwelling).
The band of heat flux density in the coastal upwelling region is visible in Fig. 3 (see also
Schmitt et al. 1989). This band is much wider than the coastal upwelling region, partially
due to the existence of an extense upward Ekman pumping region (which reaches 200-300
km from the coast in areas adjacent to capes: Bakun and Nelson 1991, McClain and
Firestone 1993) and to the existence of a wide region of enhanced mixing associated to
the coastal jet (Foo 1981, Kundu and Beardsley 1991, Pelegrf and Richman 1993). The
main reason, however, is probably related to the upwelled water moving into warmer
interior regions where it continues gaining heat. We may estimate the heat flux density
necessary to warm the upwelling water as pCp(Oo - Oi)uehl L, where ueh is the offshore
Ekman transport per meter of coastline, L is the width of the upwelling region, and Oi and
00 are the surface temperatures of the upwelling and offshore areas, respectively. Using
(}i = 18°C, (}o = 23°C, L = 250 km and ueh = 1 m 2 Is gives 80 W 1m 2 • Fig. 1 shows the
remarkable heat flux density from the atmosphere in the coastal upwelling region at 23°N,
the highest in the North Atlantic, as obtained from weather data; the annually averaged
heat influx density, as calculated from Fig. 1, is Qn = 140 W 1m 2 • Comparing these two
values we find that heat transport divergence in the coastal upwelling region is rather large,
about 60% the heat influx density. The heat transported away from the upwelling region
due to offshore Ekman transport can be estimated multiplying our previous estimate by
L which gives 1.1 x 10 13 W, about 25% the annually averaged heat influx in the AC-CC
region. During winter, however, heat influx associated to coastal upwelling is the only
mechanism that continues operating. Fig. 1 also shows, for comparison, the heat flux
densities in the trade region, slightly north of the intertropical convergence zone (ITZC);
in this case the incoming radiation flux density is also high, although not as much as in
the coastal upwelling region, but loss through latent heat flux density is comparably large.