SLEMSA
~ VARYING K
REVISED SLOPE
7. E.CANADA
USLE MODEIL_MUSLE
/
\
TASK
NOCN
Q(S’,P
S',P',CN y
USDA AMC-II CN METHOD
', A(Q,S,L,K,C,P)
Q(CN,S"),A(Q,S,L,K,C,P)
USGS LUDA CN METHOD
LANDSAT CN METHOD
USDA AMC-II CN __Q=(Q(CN,S");A(Q,...)
USGS LUDA CN ....
Q=(Q(CN,S");AQ,...)
NO LANDSAT CN Q=(Q(CN,S);A(Q,...)
USLE PRMS ALL PARAMETERS KNOWN A(R,S,K,L,C,P)
NOR SLOPEUNKNOWN ___ NO SOLUTION
—NOK
NO
NOC,P
m
S
NOL
LENGTH «-4m, S = S(SLOPE)
1.1%<SLOPE<9%,S=S'(SLOPE)
9.1%<SLOPE<30%,S=S"(SLOPE)
Figure 4: Knowledge About Soil Loss Estimation and Modelling Processes(Source,
E.G. Mtalo, 1990, page 168).
WISCHMEIER ____
HOLY PERIOD C'=15
fo USDA: PERIOD 2 C" 2 15
METHO PERIOD 3 C'=15
PERIOD 4 C'=15
C-FACTOR C PERIOD 5 C" 225
LOCATION....F REDERICTON
GRANDFALLS POTATO,GRAIN & HAY (P/P/P/G/H) __C=0.23
POTATO(PPPP) son = 0.53
GRANGGGS) D m OLD
POTATO&BROCCOLI(P/B) C = 0.56
POTATQ&GRAIN
(P/P/P/P/G) ___ C=048
(P/P/P/G) ____ C=0.40
(P/P/G)__. C=0.38
(PO) C=032
(P/G/...): Crop Rotation; e.g. Grain After Potato...
HOLY: First year corn after meadow, residue left and incorporated by ploughing.
PERIOD n: Crop stage n.
Figure 5: A Segment of Knowledge On Selection of USLE C FACTORS
systems technology. Their response was generally favorable
and they pointed out the need to carry out a more exhaustive -
investigation for the purposes of refining the prototype.
3.2 A Structured Approach to the
Data and Facts.
Compilation of
The LEARN sub-system is an expert shell which captures and
organizes knowledge in the form of a semantic network of
English-like subject-verb-object triplets. Knowledge
representation using the LEARN sub-system involves the
analysis and re-structuring of domain knowledge into Subject-
Verb-Object (SVO) triplets where, "Subject", represents a
domain concept, "Verb", represents a semantic relationship and,
"Object", represents an attribute or property of the "Subject". If
202
the "Object" part is a phrase assigning value to an attribute of the
object then the SVO triplet is equivalent to an "Object-Attribute-
Value" (OAV) triplet.
The query processor of the LEARN sub-system consists of
several specialized inference operators, collectively, referred to
as the inference layer (Figure 3). These operators exploit
semantic links among the SVO object triplets to determine class
relationships and extract the attribute values of the stored
triplets. A section of the LEARN sub-system operators and their
functional characteristics are shown in Figure 7 where S, V, O,
SV, SO, VO and SVO are basic search operators, (S,v,0)
represents the search space (database) and (S), (V), (O) and
(True, False) represent solution spaces.