BAYESIAN NETWORKS: A CONCEPTUAL FRAMEWORK FOR HABITAT RISK ASSESSMENT

cabbageswerveΤεχνίτη Νοημοσύνη και Ρομποτική

7 Νοε 2013 (πριν από 3 χρόνια και 9 μήνες)

74 εμφανίσεις

BAYESIAN NETWORKS: A CONCEPTUAL FRAMEWORK FOR
HABITAT RISK ASSESSMENT
KELLI
TAYLOR,'
College of Natural Resources, Utah State University
&
NASA
Afliliated
Research
Center, P.O. Box 1204, Logan,
UT
84322, USA
Abstract:
Landscape-scale risk assessments provide a useful tool in regional land use planning efforts.
Human related activities in particular have both direct and indirect influences on habitat quality and
wildlife viability. With limited scientific information about landscape scale effects, managers often rely
on expert knowledge to make resource appraisals. Applying Bayesian networks as
an
assessment tool
may provide land managers a method to
quantify
scientific uncertainty of outcome probabilities given a
set of land use decisions.
Bayesian networks are currently being applied to a multi-scale habitat risk assessment in the
southwestern portion of the Greater Yellowstone known as the Southern Wind River Landscape.
Bayesian Networks are utilized to assess the probabilities of risk to mule deer, sage grouse and
mink
habitats given particular anthropogenic activities including energy development and land conversion. The
intended outcome of this process is to provide a conceptual
fi-amework
for assessing risk using habitat as
a measure of habitat and land cover connectivity; whereby current knowledge and understanding of risk
are spatially represented to generate landscape level and site-specific risk maps.
WESTERNSTATES
AND
PROVmCES
DEER
AND
ELK
WORKSHOP
5:33
'
E-mail:
kjt@gis.usu.edu