Geoscience Reference
In-Depth Information
Time (CDT)
Fig. 4.10 Diurnal variation of half-hourly values of different evaporative flux ratios, obtained from the
flux data shown in Figure 4.9;
L
e
E
/α
s
R
s
(inverted open triangles);
L
e
E
/
(
R
n
−
G
) (open
triangles);
L
e
E
/
R
n
(open squares);
L
e
E
/
R
s
(open circles);
L
e
E
/
H
(
≡
Bo
−
1
, solid
squares);
L
e
E
/
G
(solid circles);
L
e
E
/
(
R
ld
−
R
ld(night)
) (solid triangles); and
L
e
E
/
(
R
lu
−
R
lu(night)
) (solid diamonds). Note that the curves of the inverted open triangles, but
especially those of the open triangles, the open squares and the open circles are nearly
coincident during the daytime hours, so that they may not be easy to distinguish. (From
Brutsaert and Sugita, 1992.)
flux (
R
n
−
G
)or(
L
e
E
+
H
), net radiation
R
n
, incoming short-wave radiation
R
s
and also,
but less so, the reflected short-wave radiation
α
s
R
s
. The fact that equilibrium evaporation
E
e
is strongly related with the available energy flux (
R
n
−
G
) (see Equation (4.30)), explains
that Equation (4.51) also works well when
F
is taken as that quantity. Note that when
F
is taken as the sensible heat flux
H
, so that ER
−
1
is the Bowen ratio Bo, self-preservation
appears to be considerably less robust than in the case of
F
=
(
R
n
−
G
). It was shown in
Crago and Brutsaert (1996) that this is caused by the different error propagation properties
of EF and Bo. Figure 4.11 also shows that self-similarity is not applicable at night. As a
further illustration, Figure 4.12 shows the evolution of the daytime evaporative fraction EF
during a long drying period over the same tallgrass prairie terrain at station 26; on any day
EF remained fairly invariant during the daytime hours, but it decreased from day to day as
the soilmoisture was gradually declining (see also Figure 2.22).
In practical applications, this concept of self-similarity is implemented as follows. If ER
is sufficiently constant during the day, the instantaneous evaporation rate at any moment
during the daytime can be estimated with (4.51), that is
L
e
E
i
=
ER
d
F
i
(4.52)
