Atlantic Coastal Biome
235
the 400 g cal cm
−1 d
−1 of boreal winter to about 250 g cal cm
−1 d
−1 in boreal summer.
Surface temperatures consequently fall to 24–25
C and the isotherms for 19–23
C become
more widely spaced. Lutjeharms and Meeuwis (1987) described as an “upwelling cell”
the region of cool surface water seen in satellite thermal imagery between Fernando Po
and the coast, mostly in March to August, during the period of greatest rainfall on the
Cameroons. This interpretation must be incorrect because these supposed “cells” occur
over an area of unusually shallow and muddy continental shelf. At this season, river
discharges are often visible in ocean color images as pigmented plumes, especially in
the Guinea–Sierra Leone and Nigeria-Cameroon sections of the coast. Thus, during the
June–September wet season, surface salinity is strongly reduced by the extremely heavy
rainfall both in the Bight of Biafra and off Sierra Leone, in both places associated with
hilly country behind the coast. During the dry, Harmattan season of NE trades (January–
March), significant falls of aeolian dust occur at sea in the western part of the region
and a secondary, evaporative cooling of sea surface temperature occurs over the whole
region. During the dry season, a sharp pycnocline lies below a shallow (30 m) mixed layer
at ∼27
C that in the wet season is almost eroded by seasonal wind stress, so that surface
temperatures fall to 24–25
C and the isotherms for 19–23
C become widely spaced.
From eastern Liberia to western Nigeria, episodic shoaling of the thermocline is sufficiently strong that it constitutes coastal upwelling. This process is highly variable between
years and is strongest along the coastlines of Ghana and the Ivory Coast, contiguous with
the “Dahomey Gap” in the forest cover ashore. Upwelling occurs especially frequently off
Cape Palmas (7
W) and Cape Three Points (2
W). Though a general cooling associated
with the shoaling and eroding of the thermocline occurs each year, significant localized
upwelling occurs only sporadically except off this central region, and even here there are
exceptional years, such as 1968, when it does not occur. Upwelling is not a continuous
process and the general season of cool water at the coast from June to September is
punctuated by a series of isolated cold events, each lasting for a week or two (Longhurst,
1964; Philander, 1979; Houghton and Colin, 1986).
The exact mechanism that causes cold water to upwell along the central part of the
coast is still quite obscure, though very much studied, and has been reviewed recently
by Ajao and Houghton (2002). I had previously suggested, along with others, that this
was a classical case of offshore Ekman flux driven by coastwise winds, but apparently we
were quite wrong: there is not, as Ajao and Houghton demonstrate, even a qualitative
relationship between wind speed or direction and the onset of upwelling on the central
coast. Several other mechanisms have been proposed, and it is now thought that a
mechanism involving coastally trapped internal waves must be involved. A fortnightly
coastal wave propagates westward at around 05 m sec
−1 , having a period exactly that
of the lunisolar tide, and this appears to have its origin on the shelf of the Niger delta,
where M 2 and S 2 tides interact in a nonlinear fashion through bottom friction. These
waves travel westward along the Guinea coast, doming the thermocline and nutricline as
they pass (especially where steep or protruding topography is encountered), forcing local
upwelling events that may cause the westward undercurrent to surface, so that the events
can be tracked westward along the coast using individual coastal station data. Though
local wind stress is usually of the correct sign to generate coastal Ekman divergence, it
is of quite insufficient strength to induce the observed vertical movement that occurs
down to 500 m at the start of the cooling season. Whether these shelf waves, or a free
equatorial wave propagating poleward along the coast, or even another form of trapped
coastal wave, are responsible for the observed upwelling remains moot at this time.
There is some evidence from satellite IRT images of sea surface temperature of shelfbreak fronts, and of shelf-break trapping of the flow from the Canary Current south of
Cape Roxo and along the edge of the Guinea shelf, especially in boreal summer. Biological
observations suggest that off both the Guinea and Nigerian shelves an amplification of
235
the 400 g cal cm
−1 d
−1 of boreal winter to about 250 g cal cm
−1 d
−1 in boreal summer.
Surface temperatures consequently fall to 24–25
C and the isotherms for 19–23
C become
more widely spaced. Lutjeharms and Meeuwis (1987) described as an “upwelling cell”
the region of cool surface water seen in satellite thermal imagery between Fernando Po
and the coast, mostly in March to August, during the period of greatest rainfall on the
Cameroons. This interpretation must be incorrect because these supposed “cells” occur
over an area of unusually shallow and muddy continental shelf. At this season, river
discharges are often visible in ocean color images as pigmented plumes, especially in
the Guinea–Sierra Leone and Nigeria-Cameroon sections of the coast. Thus, during the
June–September wet season, surface salinity is strongly reduced by the extremely heavy
rainfall both in the Bight of Biafra and off Sierra Leone, in both places associated with
hilly country behind the coast. During the dry, Harmattan season of NE trades (January–
March), significant falls of aeolian dust occur at sea in the western part of the region
and a secondary, evaporative cooling of sea surface temperature occurs over the whole
region. During the dry season, a sharp pycnocline lies below a shallow (30 m) mixed layer
at ∼27
C that in the wet season is almost eroded by seasonal wind stress, so that surface
temperatures fall to 24–25
C and the isotherms for 19–23
C become widely spaced.
From eastern Liberia to western Nigeria, episodic shoaling of the thermocline is sufficiently strong that it constitutes coastal upwelling. This process is highly variable between
years and is strongest along the coastlines of Ghana and the Ivory Coast, contiguous with
the “Dahomey Gap” in the forest cover ashore. Upwelling occurs especially frequently off
Cape Palmas (7
W) and Cape Three Points (2
W). Though a general cooling associated
with the shoaling and eroding of the thermocline occurs each year, significant localized
upwelling occurs only sporadically except off this central region, and even here there are
exceptional years, such as 1968, when it does not occur. Upwelling is not a continuous
process and the general season of cool water at the coast from June to September is
punctuated by a series of isolated cold events, each lasting for a week or two (Longhurst,
1964; Philander, 1979; Houghton and Colin, 1986).
The exact mechanism that causes cold water to upwell along the central part of the
coast is still quite obscure, though very much studied, and has been reviewed recently
by Ajao and Houghton (2002). I had previously suggested, along with others, that this
was a classical case of offshore Ekman flux driven by coastwise winds, but apparently we
were quite wrong: there is not, as Ajao and Houghton demonstrate, even a qualitative
relationship between wind speed or direction and the onset of upwelling on the central
coast. Several other mechanisms have been proposed, and it is now thought that a
mechanism involving coastally trapped internal waves must be involved. A fortnightly
coastal wave propagates westward at around 05 m sec
−1 , having a period exactly that
of the lunisolar tide, and this appears to have its origin on the shelf of the Niger delta,
where M 2 and S 2 tides interact in a nonlinear fashion through bottom friction. These
waves travel westward along the Guinea coast, doming the thermocline and nutricline as
they pass (especially where steep or protruding topography is encountered), forcing local
upwelling events that may cause the westward undercurrent to surface, so that the events
can be tracked westward along the coast using individual coastal station data. Though
local wind stress is usually of the correct sign to generate coastal Ekman divergence, it
is of quite insufficient strength to induce the observed vertical movement that occurs
down to 500 m at the start of the cooling season. Whether these shelf waves, or a free
equatorial wave propagating poleward along the coast, or even another form of trapped
coastal wave, are responsible for the observed upwelling remains moot at this time.
There is some evidence from satellite IRT images of sea surface temperature of shelfbreak fronts, and of shelf-break trapping of the flow from the Canary Current south of
Cape Roxo and along the edge of the Guinea shelf, especially in boreal summer. Biological
observations suggest that off both the Guinea and Nigerian shelves an amplification of