wave. According to Kraus and Businger (1994), boundary-layer separation
of the airflow from surface waves under this condition is rather rare event.
accompanied by a gravity-capillary wave train near the wave crest linked to
the microscale wave breaking. When wind reaches the speed at which the
surface wind drift speed exceeds the minimum phase speed of the gravitycapillary waves, a K-H instability of the gravity-capillary wave train disrupts
the air-sea interface destroying the viscous sublayer. This leads to the
separation of the airflow form waves and reduction of the drag coefficient.
The disruption of the air-sea interface also generates spume. As it has
already been pointed in Section 6.2.2, the largest droplets are produced by
the mechanism of direct splashing. The initial instability is of the K-H type.
According to Koga (1981), small projections in fact develop near the crest of
the wave, where the wind stress is generally highest, and then break up to
form droplets.
The following nondimensional number characterizes the condition for
development of the K-H instability at the air-water interface:
*
2
/
a
s
a
u
K
gV U U
.
(6.45)
The instability first occurs at
cr
K K
t
. An estimate of K cr from the
laboratory experiment of Koga (1981) is cr
K | 0.038, which at a 20
o
C air
temperature and normal atmospheric pressure corresponds to the friction
velocity in air
1.24
a
u
m s
-1 (or to friction velocity in the water
0.043
u
m s
-1 ), and the wind speed at 10 m height 10 30
U |
m s
-1 .
The spreading mechanism of the droplets varies according to the size of
the droplets. Koga and Toba (1981) concluded that for droplets with radius
of 75 Pm or more, the diffusion by air turbulence is not of great importance
(which is consistent with the graph shown in Figure 6-18). More important
is the initial speed at the instant of the droplet production and the
acceleration by wind.
Figure 6-22 shows the spatial distribution of the velocity vector of the
splashing droplets relative to the wave profile in the laboratory experiment
of Koga (1981). The experimental set up was able to resolve droplets with
radius larger than 0.4 mm. Most of these large droplets are produced on the
leading slope near the crest of wave with initial speeds of the same order, or
somewhat larger than, the speed of the projection. They spread upward and
forward by the acceleration due to the drag of the pulsating wind flowing
over wave profile. Due to gravity most of the droplets return to the water
surface on the windward slope near the crest of the next wave; their
Chapter 6: HIGH WIND SPEED REGIME
437
Banner and Phillips (1974) identified the presence of a stagnation point
of the airflow from surface waves under this condition is rather rare event.
accompanied by a gravity-capillary wave train near the wave crest linked to
the microscale wave breaking. When wind reaches the speed at which the
surface wind drift speed exceeds the minimum phase speed of the gravitycapillary waves, a K-H instability of the gravity-capillary wave train disrupts
the air-sea interface destroying the viscous sublayer. This leads to the
separation of the airflow form waves and reduction of the drag coefficient.
The disruption of the air-sea interface also generates spume. As it has
already been pointed in Section 6.2.2, the largest droplets are produced by
the mechanism of direct splashing. The initial instability is of the K-H type.
According to Koga (1981), small projections in fact develop near the crest of
the wave, where the wind stress is generally highest, and then break up to
form droplets.
The following nondimensional number characterizes the condition for
development of the K-H instability at the air-water interface:
*
2
/
a
s
a
u
K
gV U U
.
(6.45)
The instability first occurs at
cr
K K
t
. An estimate of K cr from the
laboratory experiment of Koga (1981) is cr
K | 0.038, which at a 20
o
C air
temperature and normal atmospheric pressure corresponds to the friction
velocity in air
1.24
a
u
m s
-1 (or to friction velocity in the water
0.043
u
m s
-1 ), and the wind speed at 10 m height 10 30
U |
m s
-1 .
The spreading mechanism of the droplets varies according to the size of
the droplets. Koga and Toba (1981) concluded that for droplets with radius
of 75 Pm or more, the diffusion by air turbulence is not of great importance
(which is consistent with the graph shown in Figure 6-18). More important
is the initial speed at the instant of the droplet production and the
acceleration by wind.
Figure 6-22 shows the spatial distribution of the velocity vector of the
splashing droplets relative to the wave profile in the laboratory experiment
of Koga (1981). The experimental set up was able to resolve droplets with
radius larger than 0.4 mm. Most of these large droplets are produced on the
leading slope near the crest of wave with initial speeds of the same order, or
somewhat larger than, the speed of the projection. They spread upward and
forward by the acceleration due to the drag of the pulsating wind flowing
over wave profile. Due to gravity most of the droplets return to the water
surface on the windward slope near the crest of the next wave; their
Chapter 6: HIGH WIND SPEED REGIME
437
Banner and Phillips (1974) identified the presence of a stagnation point