THE NEAR-SURFACE LAYER OF THE OCEAN
2
0
*
/
0.11 /
a
C
a
a
a
z
a u
g
u
Q
(1.42)
where D C =0.011 for the COARE algorithm, and Q D is the kinematic
molecular viscosity of the air.
The profile stability functions near neutral stratification conditions F
\ are
known from accurate measurements over land, in particular, from the Kansas
experiment (see Section 1.7.2). These functions have been extended
theoretically for highly convective conditions (Fairall et al., 1996). For
highly stable conditions, the Kansas functions predict zero fluxes. Later,
Beljaars and Holstlag (1991) found finite, but highly intermittent, values for
fluxes in very stable conditions and corrected the Kansas functions for the
stable stratification limit. Adopting the stability functions with improved
convective and stable limits eliminates occasional pathological results
obtained with earlier versions of the COARE algorithm.
The addition of gustiness provides finite scalar fluxes as the wind speed
approaches zero. In the COARE algorithm the gustiness is introduced using
the convective velocity scale, w ,
1/ 3
g
c
c
v c
g
U
w
w
z
T
E
E
§
·
c
c 4
¨
¸
©
¹
,
(1.43)
where z c is the depth of the convective boundary layer; 4 Q is the virtual
potential temperature, which is analogous to the potential temperature that
removes the adiabatic temperature variations caused by changes in the
ambient pressure of an air parcel but also accounts for humidity effects
(Stull, 1988), and c
E is an empirical coefficient (
1.25
c
E |
).
Figure 1-2 shows the comparison between measured and modeled
(COARE bulk flux algorithm) velocity transfer coefficients as a function of
atmospheric variables (U , T, q) to a bulk flux algorithm at reference height z
and the surface properties (current vector, temperature). The surface value
for specific humidity is computed from the sea surface temperature and the
vapor pressure of seawater. Strictly speaking (1.31) requires the true
interface temperature, T 0 , and salinity, S 0 , but usually only the temperature
and salinity at some depth are available. The vapor pressure of seawater is
0.98 times the vapor pressure of pure water (Kraus and Businger, 1994). The
dependence of the vapor pressure on the typical salinity changes in the nearsurface layer is negligible (except during heavy rainfalls when S 0 may drop
for as much as a few psu, see Chapters 2 and 4). When it is not raining, the
salinity diffusion sublayer implies that the true salinity is only a few tenths
psu higher than the bulk water salinity near the surface.
16
a
wind speed. Typically, measurements or model output provide input
2
0
*
/
0.11 /
a
C
a
a
a
z
a u
g
u
Q
(1.42)
where D C =0.011 for the COARE algorithm, and Q D is the kinematic
molecular viscosity of the air.
The profile stability functions near neutral stratification conditions F
\ are
known from accurate measurements over land, in particular, from the Kansas
experiment (see Section 1.7.2). These functions have been extended
theoretically for highly convective conditions (Fairall et al., 1996). For
highly stable conditions, the Kansas functions predict zero fluxes. Later,
Beljaars and Holstlag (1991) found finite, but highly intermittent, values for
fluxes in very stable conditions and corrected the Kansas functions for the
stable stratification limit. Adopting the stability functions with improved
convective and stable limits eliminates occasional pathological results
obtained with earlier versions of the COARE algorithm.
The addition of gustiness provides finite scalar fluxes as the wind speed
approaches zero. In the COARE algorithm the gustiness is introduced using
the convective velocity scale, w ,
1/ 3
g
c
c
v c
g
U
w
w
z
T
E
E
§
·
c
c 4
¨
¸
©
¹
,
(1.43)
where z c is the depth of the convective boundary layer; 4 Q is the virtual
potential temperature, which is analogous to the potential temperature that
removes the adiabatic temperature variations caused by changes in the
ambient pressure of an air parcel but also accounts for humidity effects
(Stull, 1988), and c
E is an empirical coefficient (
1.25
c
E |
).
Figure 1-2 shows the comparison between measured and modeled
(COARE bulk flux algorithm) velocity transfer coefficients as a function of
atmospheric variables (U , T, q) to a bulk flux algorithm at reference height z
and the surface properties (current vector, temperature). The surface value
for specific humidity is computed from the sea surface temperature and the
vapor pressure of seawater. Strictly speaking (1.31) requires the true
interface temperature, T 0 , and salinity, S 0 , but usually only the temperature
and salinity at some depth are available. The vapor pressure of seawater is
0.98 times the vapor pressure of pure water (Kraus and Businger, 1994). The
dependence of the vapor pressure on the typical salinity changes in the nearsurface layer is negligible (except during heavy rainfalls when S 0 may drop
for as much as a few psu, see Chapters 2 and 4). When it is not raining, the
salinity diffusion sublayer implies that the true salinity is only a few tenths
psu higher than the bulk water salinity near the surface.
16
a
wind speed. Typically, measurements or model output provide input