THE NEAR-SURFACE LAYER OF THE OCEAN
high-wind speed conditions and developed seas (Figure 3-19) that appeared
to confirm the conclusion of Stewart and Grant (1962) and Soloviev et al.
(1988) that the main part of the wave-generated turbulence dissipates within
a near-surface layer, whose depth is less than one significant wave height.
Tower-based turbulence measurements made in a lake by Kitaigorodskii
et al. (1983) and Agrawal et al. (1992) under a wide range of wind speed
conditions produced evidence in favor of a thicker layer of wave enhanced
turbulence and higher turbulence levels. Terray et al. (1996) proposed a new
scaling that accounted for the limited fetch in the lake observations and
dramatically reduced the difference between the two groups of data. Some
differences, however, could not be explained solely by difference in fetch
and wave age and might be related to methodical issues. The interpretation
of the above tower-based turbulence measurements is somewhat uncertain
because the transfer from frequency to wavenumber domain is not well
defined for oscillating flows.
For turbulence measurements, Soloviev et al. (1988) used a free-rising
profiler, Greenan et al. (2001) employed a free-gliding instrument, and
Thorpe et al. (2003a) utilized an autonomous underwater vehicle (AUV).
All of those were moving instruments: It is apparently easier to satisfy
Taylor’s hypothesis of frozen turbulence with a moving instrument. All of
those three instruments at some extent follow surface waves and, hence,
have a tendency to provide the data in the Lagrangian wave following
coordinate system (see Section 3.1.1). At the same time, the
parameterization of Terray et al. (1996) is based on tower data that are
collected and interpreted in a fixed coordinate system.
Energy budget considerations provide an estimate of the wavelength O dis
where the transition toward the dissipation regime occurs (Kitaigorodskii,
1991):
2 / 3
0
1
2
/( )
dis
E
a g
O
S
(3.82)
where E 0 is the energy flux from the region of energy input through the nondissipative region of the wave spectrum toward the dissipation subrange, and
4
1 1 10
a
| u
according to the most recent estimates of Gemmrich and
Farmer (1999). Following Gemmrich et al. (1994), we equate this energy
flux to the integral dissipation of surface wave energy due to wave breaking
in the upper ocean, which in stationary conditions is equal to the flux of the
TKE at the air-sea interface, E 0 = F 0 . According to Pierson and Moskowitz
(1964), for the saturated surface wave spectrum, the dominant wavelength
2
2
2
/
p
a u g
O
S
|
, where
5
2 8.3 10
a
u
. From equations (3.35) and (3.82), it
follows that
2 / 3
1 2
/
/(
) 0.26
dis
p
w
a a
O O D
|
. This means that breaking waves
196
high-wind speed conditions and developed seas (Figure 3-19) that appeared
to confirm the conclusion of Stewart and Grant (1962) and Soloviev et al.
(1988) that the main part of the wave-generated turbulence dissipates within
a near-surface layer, whose depth is less than one significant wave height.
Tower-based turbulence measurements made in a lake by Kitaigorodskii
et al. (1983) and Agrawal et al. (1992) under a wide range of wind speed
conditions produced evidence in favor of a thicker layer of wave enhanced
turbulence and higher turbulence levels. Terray et al. (1996) proposed a new
scaling that accounted for the limited fetch in the lake observations and
dramatically reduced the difference between the two groups of data. Some
differences, however, could not be explained solely by difference in fetch
and wave age and might be related to methodical issues. The interpretation
of the above tower-based turbulence measurements is somewhat uncertain
because the transfer from frequency to wavenumber domain is not well
defined for oscillating flows.
For turbulence measurements, Soloviev et al. (1988) used a free-rising
profiler, Greenan et al. (2001) employed a free-gliding instrument, and
Thorpe et al. (2003a) utilized an autonomous underwater vehicle (AUV).
All of those were moving instruments: It is apparently easier to satisfy
Taylor’s hypothesis of frozen turbulence with a moving instrument. All of
those three instruments at some extent follow surface waves and, hence,
have a tendency to provide the data in the Lagrangian wave following
coordinate system (see Section 3.1.1). At the same time, the
parameterization of Terray et al. (1996) is based on tower data that are
collected and interpreted in a fixed coordinate system.
Energy budget considerations provide an estimate of the wavelength O dis
where the transition toward the dissipation regime occurs (Kitaigorodskii,
1991):
2 / 3
0
1
2
/( )
dis
E
a g
O
S
(3.82)
where E 0 is the energy flux from the region of energy input through the nondissipative region of the wave spectrum toward the dissipation subrange, and
4
1 1 10
a
| u
according to the most recent estimates of Gemmrich and
Farmer (1999). Following Gemmrich et al. (1994), we equate this energy
flux to the integral dissipation of surface wave energy due to wave breaking
in the upper ocean, which in stationary conditions is equal to the flux of the
TKE at the air-sea interface, E 0 = F 0 . According to Pierson and Moskowitz
(1964), for the saturated surface wave spectrum, the dominant wavelength
2
2
2
/
p
a u g
O
S
|
, where
5
2 8.3 10
a
u
. From equations (3.35) and (3.82), it
follows that
2 / 3
1 2
/
/(
) 0.26
dis
p
w
a a
O O D
|
. This means that breaking waves
196