Simulatio
n
mode
l
(scientific
)
Im
pr
o
ve
m
o
de
l
(quality
)
Fa
rm
er
s,
Fa
rm
busines
s
advisor
sNO
P
IN
D
Interview
s
NO
P
IN
D Qu
es
tio
n
n
ai
re
–
verificatio
n
NO
P
OT
:Researcher
s
M
KA/
M
M
(re
nements
)
KA
M
Has
e
I
Simulatio
n
mode
l
(scientific
;
discussio
n
su
pp
o
rt
)
Im
pr
o
ve
m
o
de
l
(quality
,
acceptance
)
Farmers
,loca
l
authorities
,
association
s
IN
D Reques
t
fo
rdat
a
N
/media/loftslag/Hare-2011-ParticipatoryModelling.pdf
it will be one of the main aspects
of this paper.
2.3. Properties
The elements of a Fuzzy Cognitive Map are as follows:
Concepts: C1, C2, . . ., Cn. These represent the drivers and
constraints that are considered of importance to the issue under
consideration.
State vector: A = (a1, a2, . . ., an), where ai denotes the state of the
node Ci. The state vector represents the value of the concepts
/media/loftslag/Kok_JGEC658_2009.pdf
[Q(t);Q(t 1);Q(t 2)]
E[Q(ui);Q(ui 1);Q(ui 2)]
, ai =
Qb(t)
Qb(ui)
, bi =
Q(t) Qb(t)
Q(ui) Qb(ui)
,
and Qb(t) and Qb(ui) are baseflows calculated using the UKIH baseflow separation method
(Piggott et al., 2005). All rescaling coefficients were limited to a minimum value of 0.25 and a
maximum value of 5.
4.5 Deterministic prediction
A deterministic forecast was derived from the ensemble by taking
/media/vedurstofan/utgafa/skyrslur/2013/VI_2013_008.pdf