After field checking [MARSMAN and DE GRUIJTER,
1986] this model was found to provide correct
grazing suitability predictions in 95% of cases.
3.2 Soil Drainage Status
Drainage status is linked to the height of the
water table, and more particularly its Mean
Highest Water Level (or GHG value), as follows:
Drainage Status Level|GHG cm below the surface
1 »80
2 40-80
3 25-40
4 15-25
5 «15
Following field testing [MARSMAN and DE GRUIJTER,
1986] it was found that the standard deviation of
the GHG is 14cm. Using estimation by confidence
intervals the probability of land parcel vith a
certain measured GHG value being in a specified
Drainage Status Level can be calculated. For
example with a GHG value of 60cm, the probability
of the parcel being in Drainage Status Level 2 is
85%.
3.3 Soil Bearing Capacity
Bearing capacity (3 classes) is related to
Soiltype (5 classes) and GHG, as follows:
Soiltype ] 2 3 4 5
GHG(cm)
0-12 3- 3-3 3-3
13-24 3:99: 9:72
235-33 3 22 3 2
34-40 2. à 3 2 |
41-60 2 2.2.2 1
61-80 1 12 22
80-140 1:1 T2 1
Thus, e.g., Soiltype 3 with a water table 41-60 cm
below the surface has a Bearing Capacity Class of
2,
In its turn Soiltype is related to Soiltexture as
follows:
Soiltype Organic | Clay
content | content
1. Peat 15-100% 0-8%
2. Clay with peat underlay | 22-70% 8-100%
3. Clay 0-15% 25-100%
4. Clayey sand 0-2.5% 8-25%
5. Sand 0-2.5% 0-8%
As bearing capacity is determined from GHG,
organic content, and clay content, the qualities
of all three need to be known. Tests have shown
that the probability of these particular organic
content and clay content classes being correct is
98% [MARSMAN and DE GRUIJTER, 1986]. The quality of
GHG data was discussed in the previous section,
and an example landparcel was shown to have a
probability of 85% that it was in its stated
Drainage Status Level (or GHG level). Thus in the
same example landparcel, the probability of its
Bearing Capacity Class being correct (Pbc) is:
360
Pbc = 0.85 x 0.98 x 0.98 = 0.82 = 82%
3.4 Soil Moisture Supply Capacity
Moisture supply capacity is recorded in
millimeters and is calculated using a polynomial
of twenty coefficients and three variables
(rooting depth, mean lowest water-table depth, and
mean spring water-table depth) [RAMLAL, 1991]. In
this application it is reclassified into 5
discrete classes:
Moisture Supply
Capacity Class
Moisture Supply
Capacity (mm)
>200
150-200
100-150
50-100
<50
LW =
Following error propagations carried out by the
Dutch Soil Research Institute [MARSMAN and DE
GRUIJTER, 1986] it was found that the standard
deviation of Moisture Supply determinations is
17mm. With this information, and using estimation
by confidence intervals the pro- bability (e.g.)
of a landparcel having Moisture Supply Capacity
Class 2, when its Moisture Supply Capacity has
been measured to be 166mm, is 81%.
3.5 Quality of the
Classification
Grazing Suitability
Taking into account the quality of the model (see
section 3.1), the quality of the Soil Drainage
Status Level (section 3.2), the Soil Bearing
Capacity Class (section 3.3), the Moisture Supply
Capacity Class (section 3.4), and using Crisp Set
Theory it is possible to estimate the probability
(P) of the given landparcel (referred to in
sections 3.2, 3.3, 3.4) having the predicted
Grazing Suitability to be:
P =:0.98(0-85 x 0.82 » 0.81) = .55 = 55%
Applying Fuzzy Sub-Set Theory and using these
probabilities as Certainty Factors, the overall
Certainty Factor associated with the predicted
Grazing Suitability is 0.81, i.e. MIN(0.98,
0.85,0.82,0.81).
4. RESULTS OF EXPLORATION OF PROPOSAL
In this study a database was built which held Soil
Polygons supplied by the Dutch Soil Research
Institute, the land parcel boundaries supplied by
the Dutch Topographic Service, and database tables
holding the soil characteristics and the relevant
soil characteristics quality parameters of the
those soil polygons.
Using the existing facilities of ILWIS the Grazing
Suitability Model was inserted and a multicoloured
5-class grazing suitability map produced. Then
using the procedures outlined in Section 3 and
implemented in ILVIS the quality parameters were
processed to give i) a 2-class probability map
(«502 probability, >50% probability); ii) a
3-class probability map (low, average, and good
probability); and iii) a 5-class probability map
(<10%, 10-30%, 30-40%, 40-50%, and 50-60%). The
5-class map is shown below, in FIGURE 2.
The
3-cl
a ma
to
land
(bas
over
dark
(See
DRUM
for
visu:
qual:
In
unce!
that
Posi!
direi
indi
Comp:
link
worl
parar
datal
Teco!
Line:
Repo:
been
be u:
parar
As 1
proce
ask
for