Linking Threats to Assets in Complex Ecological and Socio-Economic Systems: Qualitative Modelling for Tourism Development in North Western Australia Jeffrey Dambacher & Keith Hayes CSIOR Mathematical and Information Sciences Geoff Hosack Oregon State University
25
Embed
Jeffrey Dambacher & Keith Hayes CSIOR Mathematical and Information Sciences
Linking Threats to Assets in Complex Ecological and Socio-Economic Systems: Qualitative Modelling for Tourism Development in North Western Australia. Jeffrey Dambacher & Keith Hayes CSIOR Mathematical and Information Sciences. Geoff Hosack Oregon State University. - PowerPoint PPT Presentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Linking Threats to Assets in Complex Ecological and Socio-Economic Systems: Qualitative Modelling for Tourism Development in North Western Australia
Jeffrey Dambacher & Keith HayesCSIOR Mathematical and Information Sciences
Geoff HosackOregon State University
Risk-Based Management of Natural Resources
• Depends on models• Risk assessment predicated on model linking threats to asset
• Natural systems are complex• Realistic representation of causality is difficult• Ecological and socioeconomic systems have feedback
• Experts versus stakeholder participation• Stakeholders typically not involved in model development yet live with the risk
• Model uncertainty• Difficult to address
• Results conditional on all parameters• Typically a narrow field of models considered
Threat AssetModel
Model Uncertainty
Parametric Uncertainty• Precise measurements• Expert opinion• Simulations with plausible parameter space• Receives majority of attention and effort in modelling
Model Structure Uncertainty• Within a model: feedback cycles with opposing sign• Between models: different interactions or variables• Largely ignored
“Model structure uncertainty is the 800 pound gorilla in
the middle of the room that no-one talks about”
Scott Ferson
Methods
• Causal Graphical Models
• Bayesian belief network (BBN)
• Qualitative model (QM)
• Model uncertainty
• Qualitative Prediction weights
• Merging of BBN and QM
Bayesian Belief Networks
• group of nodes connected by directed arrows such that there are no cycles (loops)
• “child” nodes with incoming arrows are probabilistically dependent on “parents” values
Impact to CoralImpact to Coral from Input to Algae from Input to Algae 358 feedback cycles 358 feedback cycles + 82, - 276, 194 net + 82, - 276, 194 net Prediction weight Prediction weight W = 194/276 = 0.54 W = 194/276 = 0.54
Negative response Negative response in coral seems in coral seems likely....likely....but how likely?but how likely?
c(seq(0, 1, 0.2))
c(se
q(0
, 2.5
, 0.5
))
0.5
1.5
2.5
Distribution(a)
Weights
Pro
ba
bili
ty o
f Co
rre
ct S
ign
0
0.2
0.4
0.6
0.8
1.0
Independence(b)
Weights
Pro
ba
bili
ty o
f Co
rre
ct S
ign (c) Dependence
c(seq(0, 1, 0.2))
c(se
q(0
, 2.5
, 0.5
))
0.5
1.5
2.5
Weights
Pro
ba
bili
ty o
f Co
rre
ct S
ign
0
0.2
0.4
0.6
0.8
1.0
Weights
Pro
ba
bili
ty o
f Co
rre
ct S
ign
c(seq(0, 1, 0.2))
c(se
q(0
, 2.5
, 0.5
))
0.5
1.5
2.5
Re
lativ
e F
req
ue
ncy
Weights
Pro
ba
biM
Ity o
f Co
rre
ct S
ign
0
0.2
0.4
0.6
0.8
1.0P
rop
ort
ion
Co
rre
ct S
ign
of a
dj(
A
)ij
Weights
Pro
ba
bili
ty o
f Co
rre
ct S
ign
c(se
q(0
, 2.5
, 0.5
))
0 0.2 0.4 0.6 0.8 1
0.5
1.5
2.5
Magnitude#a i j
Pro
ba
biR
Ity o
f Co
rre
ct S
ign
0 0.2 0.4 0.6 0.8 1
0
0.2
0.4
0.6
0.8
1.0
Pro
ba
bili
ty o
f Co
rre
ct S
ign
0 0.2 0.4 0.6 0.8 1
Prediction Weight Wi j
Testing prediction weights (W) through simulations
• Proportion of simulations with correct sign given by least square fit to non-linear function
• Sign of each element of adj(–ºA) converted into a probability and incorporated into the CPTs of a BBN via a linear relationship
Qualitative predictions to CPT
1] [0.5, range the on
functionlogistic prediction
vequantitati
prediction
equalitativ
ijijWTijW
ijijWTijWij
#ij ΤWβWβ
ΤWβWβAadjAadj
exp1
expsgnsgnPr
1][1/3, range
ony Probabilitchange no 0
decrease ve-increase ve
adj
3
1,4)(sgn(Pr ijijij
f TWA
Snowshoe hare example
Model A Model B
Sym_Adj (-°A) Sym_Adj (-°B)
Adj (-°A) W Adj (-°B) W
The null model = fully connected community matrix
V H
L
a ,2 2 a ,3 3 a ,2 3 a ,3 2 a ,1 2 a ,3 3 a ,1 3 a ,3 2
a ,1 2 a ,2 3 a ,1 3 a ,2 2
a ,2 1 a ,3 3 a ,2 3 a ,3 1 a ,1 1 a ,3 3 a ,1 3 a ,3 1
a ,1 1 a ,2 3 a ,1 3 a ,2 1
a ,2 1 a ,3 2 a ,2 2 a ,3 1 a ,1 1 a ,3 2 a ,1 2 a ,3 1