Geoscience Reference
In-Depth Information
start of ponding, as described in Section 9.4.3. This means that the evaporation rate at the
beginning of its desorptive phase is assumed to depend mostly on the water left in the soil
profile and much less on the prior evaporation history. This assumption is consistent with the
observations of Gardner and Hillel (1962) under constant laboratory conditions, mentioned
in Section 9.6.2.
The effect of vegetation
The two-stage concept with the desorption approach for the second stage was originally
developed for evaporation from bare soil. However, it was concluded by Brutsaert and
Chen (1995) that the desorption formulation can also be put to use in the description of
daily evaporation from grassland and other similar types of short vegetation. From their
analysis of experimental data obtained during several drying episodes in a natural tallgrass
prairie area in Kansas during FIFE, the First ISLSCP Field Experiment (see Sellers
et al.
,
1992), the following sequence of events appears to take place. Initially after rain, both soil
surface and vegetation evaporate at a rate governed by the available supply only; this can be
considered as a first stage of drying. (In this particular experimental setting, the first-stage
behavior lasted as long as the volumetric soil moisture
θ
in the top 10 cm was in excess
of 27%.) As the surface soil moisture becomes depleted below this first critical level, the
water supply rate to the soil surface becomes a limiting factor, but at first the plant roots are
still able to extract water from the soil at the energy-limiting rate. This may be referred to
as the transition stage. In this stage the combined rate of evaporation from the surface and
from the vegetation continues to decrease, until a second critical state of soil moisture is
reached. (In the experiment this second critical state was reached when the moisture content
went below about 17% and the vertical gradient started to exceed about 1.15% cm
−
1
at
5 cm.) At this point the vegetation becomes so stressed and relatively inactive that the
drying takes place mainly from the soil surface; from that point on the daily evaporation
proceeds like in a second stage of drying and it can be described simply by a desorption
formulation, that is proportional to the square root of time. The transition period can last
from several days to two weeks, depending on the soil moisture conditions and on the
season. The longer transition periods were observed under conditions of lower net radiation
and of higher soil moisture content at depths in excess of 50 cm. These results were in
contrast with the observations of relatively abrupt transitions for bare soil.
Example 9.4. A monthlong drought period in tallgrass prairie
The longest documented drying episode during the FIFE experiment occurred in the fall
of 1987. After a major rainfall event on day 253 (September 10) and minor rainfalls on
days 258 and 259, no rain fell until day 288 (October 15). (See also Figures 2.22 and 4.12.)
The recorded daily evaporation remained larger than the equilibrium evaporation
E
e
(see
Equation (4.30)) until day 258, when the near-surface soil moisture content
θ
was 0.303.
The evaporation rate became equal to
E
e
on day 259 and it dropped below it after that;
therefore, day 260 was taken to be the start of the transition stage. In order to determine
the end of the transition, the data were then analyzed as suggested by Equation (9.110),
and linear regression yielded, as shown in Figure 9.27, (
L
e
E
)
−
2
=
2
.
0 (
t
−
271)
×
10
−
4
,in
which (
L
e
E
) is the average daily latent heat flux in W m
−
2
and
t
the time as day of the year.
By choosing day 273 as the starting point for the integration, the cumulative evaporation
after the onset of the second stage could be obtained by Equation (9.111) (or (9.113)), as
shown in Figure 9.28. The fact that the data could be fitted to a straight line supports the
