Stratification and Irradiance: The Consequences of Latitude
57
induces a cascade to smaller scales and to ever-smaller eddying and, hence, mixing. This
process becomes less energetic as a function of latitude, with a slower cascade toward
the equator. Thus, wind-induced turbulence and mixing is—other things being equal—
a direct function of latitude and another reason for the physical stability of tropical
oceans.
The final consequence of latitudinal change in the Coriolis parameter is that the values
taken by the Rossby radii of deformation are latitude-dependent (Emery et al., 1984;
Houry et al., 1987). The external radius (Re) determines the length scales for barotropic
(geostrophically balanced) phenomena; it is the length scale over which gravitational
phenomena balance the tendency of f to deform a surface and may be computed as
(gH/f), where H is bottom depth and g is acceleration due to gravity. The internal
radius (Ri) also depends on stratification and determines the horizontal dimensions of
quasigeostrophic features—mesoscale eddies and Rossby waves. It may be approximated
as (1/first eigenvalue N) × f and, because f vanishes at the equator, Ri here takes a value
of infinity. At 50
latitude, Ri takes values of only 25 km whereas at 5
latitude, values
are around 300 km. For this reason, we may expect larger and fewer mesoscale eddies as
we approach the equator.
There are also latitudinal differences in the density gradients (and thus in their resistance to wind mixing) between the summer pycnocline of midlatitudes and the permanent
shallow pycnocline of low latitudes; the latter is steeper and has greater resistance to
mixing, thus enhancing the effects discussed earlier. Furthermore, the heat balance at the
surface of low-latitude oceans is such that there is a mean downward heat flux across the
sea surface that is balanced by horizontal transport to higher latitudes; seasonal changes
in incoming irradiance here are sufficiently small that at no season is there sufficient loss
of heat to produce convective deepening of the permanent tropical thermocline. Finally,
there is often an excess of precipitation over evaporation at the surface of the tropical
ocean, so mixed-layer water at low latitudes is characteristically not only warmer but also
a little fresher than deeper water. The tropical Ekman layer, therefore, lies above a pycnocline having greater stability and resistance to mixing than elsewhere; this is expressed
as a relatively high subsurface maximum of the Brunt-Väisälä buoyancy frequency (N, in
cycles hr
−1 ; see later discussion).
Finally, consider also the effect of the seasonal increase in the westward stress of the
trade winds across the tropical ocean that occurs within about 15–20
latitude of the
equator. It is this stress that maintains an upward slope of the sea surface and downward
slope of the pycnocline to the west in each ocean. The adjustment time of an ocean basin
to wind stress is related to the time taken for planetary waves to propagate across the
ocean, so an equilibrium seasonal response cannot occur within a single season across
the great width of the tropical Pacific (15,000 km). Additionally, over this great distance,
wind stress in the west is not in phase with that in the east, so the response of the ocean
cannot be simple; consequently, it is only when wind stress changes for longer periods,
on the interannual El Niño–Southern Oscillation scale, that an equilibrium response in
Pacific circulation can occur. However, the tropical Atlantic (just 5000 km wide at the
equator) responds to the seasonal cycle of trade wind stress with a seasonal basin-wide
geostrophic adjustment of mixed-layer depth that maintains equilibrium.
The aspects of ocean physics discussed in this section serve to distinguish very clearly
the characteristics of low- and high-latitude oceans as biological habitats. But they do
not suggest the discontinuities we might like to have in order to partition the apparent
continuum forced by the continuous vanishing of the Coriolis parameter equatorward.
However, if we look more closely at one feature from the previous discussion—relative
pycnocline stability, the property that most closely controls resistance of the water
column to vertical mixing—then I believe that the required discontinuities can be
demonstrated.
57
induces a cascade to smaller scales and to ever-smaller eddying and, hence, mixing. This
process becomes less energetic as a function of latitude, with a slower cascade toward
the equator. Thus, wind-induced turbulence and mixing is—other things being equal—
a direct function of latitude and another reason for the physical stability of tropical
oceans.
The final consequence of latitudinal change in the Coriolis parameter is that the values
taken by the Rossby radii of deformation are latitude-dependent (Emery et al., 1984;
Houry et al., 1987). The external radius (Re) determines the length scales for barotropic
(geostrophically balanced) phenomena; it is the length scale over which gravitational
phenomena balance the tendency of f to deform a surface and may be computed as
(gH/f), where H is bottom depth and g is acceleration due to gravity. The internal
radius (Ri) also depends on stratification and determines the horizontal dimensions of
quasigeostrophic features—mesoscale eddies and Rossby waves. It may be approximated
as (1/first eigenvalue N) × f and, because f vanishes at the equator, Ri here takes a value
of infinity. At 50
latitude, Ri takes values of only 25 km whereas at 5
latitude, values
are around 300 km. For this reason, we may expect larger and fewer mesoscale eddies as
we approach the equator.
There are also latitudinal differences in the density gradients (and thus in their resistance to wind mixing) between the summer pycnocline of midlatitudes and the permanent
shallow pycnocline of low latitudes; the latter is steeper and has greater resistance to
mixing, thus enhancing the effects discussed earlier. Furthermore, the heat balance at the
surface of low-latitude oceans is such that there is a mean downward heat flux across the
sea surface that is balanced by horizontal transport to higher latitudes; seasonal changes
in incoming irradiance here are sufficiently small that at no season is there sufficient loss
of heat to produce convective deepening of the permanent tropical thermocline. Finally,
there is often an excess of precipitation over evaporation at the surface of the tropical
ocean, so mixed-layer water at low latitudes is characteristically not only warmer but also
a little fresher than deeper water. The tropical Ekman layer, therefore, lies above a pycnocline having greater stability and resistance to mixing than elsewhere; this is expressed
as a relatively high subsurface maximum of the Brunt-Väisälä buoyancy frequency (N, in
cycles hr
−1 ; see later discussion).
Finally, consider also the effect of the seasonal increase in the westward stress of the
trade winds across the tropical ocean that occurs within about 15–20
latitude of the
equator. It is this stress that maintains an upward slope of the sea surface and downward
slope of the pycnocline to the west in each ocean. The adjustment time of an ocean basin
to wind stress is related to the time taken for planetary waves to propagate across the
ocean, so an equilibrium seasonal response cannot occur within a single season across
the great width of the tropical Pacific (15,000 km). Additionally, over this great distance,
wind stress in the west is not in phase with that in the east, so the response of the ocean
cannot be simple; consequently, it is only when wind stress changes for longer periods,
on the interannual El Niño–Southern Oscillation scale, that an equilibrium response in
Pacific circulation can occur. However, the tropical Atlantic (just 5000 km wide at the
equator) responds to the seasonal cycle of trade wind stress with a seasonal basin-wide
geostrophic adjustment of mixed-layer depth that maintains equilibrium.
The aspects of ocean physics discussed in this section serve to distinguish very clearly
the characteristics of low- and high-latitude oceans as biological habitats. But they do
not suggest the discontinuities we might like to have in order to partition the apparent
continuum forced by the continuous vanishing of the Coriolis parameter equatorward.
However, if we look more closely at one feature from the previous discussion—relative
pycnocline stability, the property that most closely controls resistance of the water
column to vertical mixing—then I believe that the required discontinuities can be
demonstrated.