Chapter 5. SPATIALLY-COHERENT STRUCTURES
At the intersection of a particular frontal interface we do not know the
intersection angle. Soloviev and Lukas (1997b) nevertheless showed that in
a statistical sense for the frontal interfaces detected with their algorithm, the
average wind-to-front angle,
E
T
, E is the wind direction (we use the
meteorological convention), and J is ҏ the ship’s course. The front vector is
defined here as a vector that is normal to the frontal line. The vector
direction is positive from less to more dense water (i.e., coinciding with the
direction of the gravitational spreading of the front).
According to Figure 5-31, the density and salinity interfaces (which are
mostly not compensated fronts) appear to depend on the wind-to-front angle.
The spatial anisotropy is pronounced for the density interfaces (Figure
5-31a), is less pronounced (but still statistically significant) for the salinity
interfaces (Figure 5-31c), and is practically non-resolvable for the
temperature interfaces (Figure 5-31e).
It is remarkable that the density fronts do not seem to depend much on
the wind speed magnitude (Figure 5-31b), while the salinity and temperature
interfaces degrade with increasing wind speed. Larger error bars in Figure
5-31b and Figure 5-31f at 14 m s
-1 average wind speed are primarily due to a
relatively small number of observations of the density and temperature fronts
under high wind speed conditions (10 and 3 respectively). For the salinity
interfaces, this number is larger (37), and the confidence limits for the 14 m
s
-1 bin are smaller (Figure 5-31d).
Based on the theory developed in Section 5.4.4, the fronts that do not
satisfy inequality (5.42) (i.e., cannot interact with wind stress) are removed
from the ensemble averaging in Figure 5-31. Removal of this relatively
small number of interfaces (see histogram in Figure 5-37b) results in some
reduction of error bars but does not modify the main conclusions about
directional anisotropy and wind speed dependence of the sharp frontal
interfaces.
5.4.3 Internal wave–shear flow interaction as a cause of repeating
frontal interfaces
Soloviev and Lukas (1997b) hypothesized that the sharp frontal
interfaces may occur as a result of nonlinear buoyant adjustment of the
stably stratified near-surface layer of the ocean to external forcing. The
external forcing includes the variable buoyancy flux and wind stress at the
ocean-air interface, tidal motions, and so forth. Relatively small thickness of
the near-surface anomalies is favorable for the development of nonlinear
interactions. The frontal interfaces are often found in groups (as in the
example shown in Figure 5-23). This feature of the frontal interfaces was
331
At the intersection of a particular frontal interface we do not know the
intersection angle. Soloviev and Lukas (1997b) nevertheless showed that in
a statistical sense for the frontal interfaces detected with their algorithm, the
average wind-to-front angle,
E
T
, E is the wind direction (we use the
meteorological convention), and J is ҏ the ship’s course. The front vector is
defined here as a vector that is normal to the frontal line. The vector
direction is positive from less to more dense water (i.e., coinciding with the
direction of the gravitational spreading of the front).
According to Figure 5-31, the density and salinity interfaces (which are
mostly not compensated fronts) appear to depend on the wind-to-front angle.
The spatial anisotropy is pronounced for the density interfaces (Figure
5-31a), is less pronounced (but still statistically significant) for the salinity
interfaces (Figure 5-31c), and is practically non-resolvable for the
temperature interfaces (Figure 5-31e).
It is remarkable that the density fronts do not seem to depend much on
the wind speed magnitude (Figure 5-31b), while the salinity and temperature
interfaces degrade with increasing wind speed. Larger error bars in Figure
5-31b and Figure 5-31f at 14 m s
-1 average wind speed are primarily due to a
relatively small number of observations of the density and temperature fronts
under high wind speed conditions (10 and 3 respectively). For the salinity
interfaces, this number is larger (37), and the confidence limits for the 14 m
s
-1 bin are smaller (Figure 5-31d).
Based on the theory developed in Section 5.4.4, the fronts that do not
satisfy inequality (5.42) (i.e., cannot interact with wind stress) are removed
from the ensemble averaging in Figure 5-31. Removal of this relatively
small number of interfaces (see histogram in Figure 5-37b) results in some
reduction of error bars but does not modify the main conclusions about
directional anisotropy and wind speed dependence of the sharp frontal
interfaces.
5.4.3 Internal wave–shear flow interaction as a cause of repeating
frontal interfaces
Soloviev and Lukas (1997b) hypothesized that the sharp frontal
interfaces may occur as a result of nonlinear buoyant adjustment of the
stably stratified near-surface layer of the ocean to external forcing. The
external forcing includes the variable buoyancy flux and wind stress at the
ocean-air interface, tidal motions, and so forth. Relatively small thickness of
the near-surface anomalies is favorable for the development of nonlinear
interactions. The frontal interfaces are often found in groups (as in the
example shown in Figure 5-23). This feature of the frontal interfaces was
331