Chapter 2: SEA SURFACE MICROLAYER
therefore revealed more detail (Figure 2-6b). According to estimates by
Soloviev and Vershinsky (1982), in the nighttime convective mixing regime
(no precipitation or insolation) the conductivity profiles in the near-surface
layer of the ocean mainly depend on the temperature rather than salinity
variations. Frictional scales of the turbulent temperature and salinity
fluctuations are
0
p
Q
T
c
u
UN
and
0
E
Q S
S
u L
UN
,
(2.6)
respectively, S 0 is the average surface salinity, L is the latent heat of
vaporization, Q E is the latent heat flux, p
c is the specific heat capacity of
water, and N is the von Karman constant.
For inhomogeneities exceeding the Kolmogorov internal length scale of
turbulence (2.2), the ratio of the temperature and salinity scales expressed in
terms of the equivalent conductivity changes is as follows:
0
0
25
T
T
T
S
S
S
p
E
LQ
C
T
C
S
c S Q
J
J
J
J
'
|
'
,
(2.7)
where
,
/
T
S p
C T
J
w w
, and
,
/
S
T p
C S
J
w w
. The estimates of the
Kolmogorov length scale for temperature (2.2) and salinity (2.3) are
0.7
T
K |
mm and D
K | 0.07 mm respectively. These estimates are made for
the conditions of experiments reported by Soloviev and Vershinsky (1982)
under an assumption that the turbulence is driven by convective instability.
Since the shear and surface wave instability can only add to the turbulence
dissipation level, these are the upper bound estimates of T
K and D
K .
According to (2.7), the contribution of temperature to the conductivity
changes during nighttime convection well exceeds that of salinity. The highresolution conductivity profiles can therefore be interpreted in terms of
temperature.
Figure 2-7 shows a series of conductivity profiles in the depth range
from 2 mm to 20 cm obtained during nighttime. The time interval between
successive profiler was from 5 to 9 minutes; the ship drifted for tens of
meters.
For the conditions of this experiment, an estimate for the flux Rayleigh
number defined according to Foster (1971) is
4
2
1 2
0
/(
) 10
f
T
T
Ra
gQ h
D
N Q |
,
where h is the mixed layer depth (equal to 50 m in this estimate). Free
convection at very large Rayleigh numbers is intermittent in space and time
(Turner, 1973). Howard (1966) formulated a phenomenological theory of the
79
therefore revealed more detail (Figure 2-6b). According to estimates by
Soloviev and Vershinsky (1982), in the nighttime convective mixing regime
(no precipitation or insolation) the conductivity profiles in the near-surface
layer of the ocean mainly depend on the temperature rather than salinity
variations. Frictional scales of the turbulent temperature and salinity
fluctuations are
0
p
Q
T
c
u
UN
and
0
E
Q S
S
u L
UN
,
(2.6)
respectively, S 0 is the average surface salinity, L is the latent heat of
vaporization, Q E is the latent heat flux, p
c is the specific heat capacity of
water, and N is the von Karman constant.
For inhomogeneities exceeding the Kolmogorov internal length scale of
turbulence (2.2), the ratio of the temperature and salinity scales expressed in
terms of the equivalent conductivity changes is as follows:
0
0
25
T
T
T
S
S
S
p
E
LQ
C
T
C
S
c S Q
J
J
J
J
'
|
'
,
(2.7)
where
,
/
T
S p
C T
J
w w
, and
,
/
S
T p
C S
J
w w
. The estimates of the
Kolmogorov length scale for temperature (2.2) and salinity (2.3) are
0.7
T
K |
mm and D
K | 0.07 mm respectively. These estimates are made for
the conditions of experiments reported by Soloviev and Vershinsky (1982)
under an assumption that the turbulence is driven by convective instability.
Since the shear and surface wave instability can only add to the turbulence
dissipation level, these are the upper bound estimates of T
K and D
K .
According to (2.7), the contribution of temperature to the conductivity
changes during nighttime convection well exceeds that of salinity. The highresolution conductivity profiles can therefore be interpreted in terms of
temperature.
Figure 2-7 shows a series of conductivity profiles in the depth range
from 2 mm to 20 cm obtained during nighttime. The time interval between
successive profiler was from 5 to 9 minutes; the ship drifted for tens of
meters.
For the conditions of this experiment, an estimate for the flux Rayleigh
number defined according to Foster (1971) is
4
2
1 2
0
/(
) 10
f
T
T
Ra
gQ h
D
N Q |
,
where h is the mixed layer depth (equal to 50 m in this estimate). Free
convection at very large Rayleigh numbers is intermittent in space and time
(Turner, 1973). Howard (1966) formulated a phenomenological theory of the
79