Geoscience Reference
In-Depth Information
The Rutter model also takes account of evaporation from the two stores based on a potential
evapotranspiration rate for a wet canopy calculated using the Penman-Monteith equation (see
Box 3.1) with a canopy resistance of zero. If either store is greater than its storage capacity, then
evaporation takes place at the potential rate. If either storage is below its minimum capacity
then it is assumed that part of the canopy is dry and evaporation is reduced proportionally as
E
a
=
E
p
S
TF
C
TF
(B3.2.2)
where
E
a
is the actual evaporation rate and
E
p
is the estimated potential rate from the inter-
ception store. A similar form is used for the stemflow store, allowing that this may account
for a proportion of the total potential evaporation (usually taken as equal to
p
t
) until that store
is dry.
The Rutter model therefore requires the specification of six parameters,
p
,
p
t
,
D
s
,
b
,
C
TF
and
C
SF
together with estimates of rainfalls and potential evaporation as inputs. In many veg-
etation types, notably crops and deciduous trees, values of the parameters are not constant
throughout the year but change with the pattern of leaf development. Care should also be
taken with adopting values found in the literature, since the values quoted may depend on
the particular form of the model used. Values for different types of tree canopy are given in
Calder (1977); Gash and Morton (1978); Dolman (1987); Gash
et al.
(1980); and Lloyd
et
al.
(1988). The most important parameter is generally the total storage capacity,
C
TF
+
C
SF
,
as this has dominant control of the amount of evaporation that takes place at the higher wet
canopy rates.
The assumptions of the Rutter model may be summarised as follows:
A1 Rainfall inputs can be apportioned between direct throughfall, canopy interception and
stemflow storage.
A2 Drainage from the canopy interception store is an exponential function of storage in excess
of a minimum storage capacity, once that capacity is exceeded.
A3 Drainage from the stemflow store takes place as soon as the storage capacity is filled.
A4 Evaporation from the canopy and stemflow stores is at the wet canopy potential rate if
storage is greater than the minimum capacity, but is linearly reduced below the potential
rate if the canopy or stems are partially wet. Note that, as the canopy dries, if the evapo-
ration from the canopy is not sufficient to satisfy the potential rate, then it is possible that
there will be some additional transport of water to the atmosphere due to transpiration or
soil evaporation).
A simplified analytical variant of the Rutter model was proposed by John Gash (1979) and has
met with reasonable success in a number of different environments (Gash
et al.
, 1980; Lloyd
et al.
, 1988; Navar and Bryan, 1994).
B3.2.3 The Calder Stochastic Model of Interception
The way in which net throughfall is related to canopy storage in the original Rutter model (which
allowed drainage even for a canopy storage less than the capacity value) means that the canopy
reaches maximum storage only after much more rainfall has fallen than the capacity volume,
while the exponential drainage function predicts a small amount of throughfall even when the
storage is zero. To avoid this problem, Ian Calder (1986) proposed an alternative stochastic
model of interception based on the probabilities of raindrops striking elemental areas making
up canopy surfaces. In the original formulation, a Poisson distribution was assumed to relate
the mean number of drops retained to the mean number of drops striking the element. The
result was a model in which interception was dependent on both storm volume and the mean
drop size of the rainfall with the effective canopy storage also being a function of drop size.
