development of a gis-based model for quantification of ... · development of a gis-based model for...
TRANSCRIPT
Development of a GIS-Based Model for Quantification of Scenic Resources
Doug PierceyESRI International User Conference
August 5, 2008
Presentation Outline
ContextVisual Quality ConceptsModel Development/DescriptionValidationUtilityNext Steps
Context
Increasing demands on the landbase– Importance of
non-traditional usesHolistic management required– Maintain sustainability– Trade-offs often
necessary
Context
Holistic management occurs– across multiple
temporal and spatial scales
– from multiple perspectives
Difficult to manage for non-traditional values– social perspective
(eg. visual quality)
Social
Economic
Ecological
Past
Present
Future
Loca
l
Reg
iona
l
Nat
iona
l
Inte
rnat
iona
l
Context
Why measure visual quality?– Indicator or catalyst for other non-traditional land
uses (eg. tourism, outdoor recreation).– It is what people see.
Visual Quality Concepts
Mapping visual quality
Local scale– 3-D rendered
landscapes– High level of detail over
small areas– Exploring options
Landscape scale– 2-D numerical
representation– Less detail over large
areas– Planning
Visual Quality Concepts
How do we quantify perception of visual quality?
711
5
104 3
1
89
Need to define a conversion between landscape components and public perception of visual quality.Visual Quality Index (VQI)– Manual exercise in the past
(sometimes subjective)– Computer-based model
desired (using GIS/RS data).
2
Model Development
Validation
LiteratureReview VQI Model
DataCollectionSurvey
Model Description
7 Landscape Components– Relative Relief– % Water– Dominant Water Type– Vegetative Variety– Topographic Variety– Degree of Alteration– Type of Alteration
All components GIS-availableHexagonal polygon mapping unit
Model Description
VisualQuality Model
• Relative Relief
•Topographic Variety
• Hexagonal polygon mapping unit
• Water Influence
F eature toR aster
S tra tumGri d
Foc al V arie ty3 X 3
S tra tVa rTemp
Re cla ss ify S t ra tMas k
St ra tumFie l d
F eature toRas te r (2)
P ol ygo nNumb er
Ras te r ToO the r Format
(mul tip le)
Deriv edWork sp ac
InputImageBands
G ri dsWork sp ace
Combi ne TMB and s
Outp ut T MBa nds
Greater T ha n Imag eA re a Mas k
Cons tantVa lue o f 0
Di sturba nc eMas k
D is t Mas k1
D is t Mas k2
D is t Mas k3
Di sturbance G ri ds
photo yr
Imag eY ear
S ing le OutputMap A lgebra A llMas k s
Sa mp le pres ampl es.dbf
CombineV egetated
Cl us ters
O utputVe gClus te rs
Ras te r ToO the r Format(mult ip l e) (2)
D eriv edWo rk s pac
Input V egCl us ters
Greater T ha n(2)
ClusterMask
R ecl ass if y (2) S ampleMask
P oly gon #
Cons ta ntVa lue o f 0
Fores tInvento ry
Ras te r ToO the r Format(mult ip l e) (3)
Deriv edWork sp ac
InputEO SD
EO SD Fore stMa sk
S ing le OutputMa p A lgeb ra
(2)
for_c lusters
OutputForestMask
• Spatial representation of visual quality
• Vegetation
Model Description
% Water Value
0 – 5 0
6 – 30 1
30 2
AlterationLevel
Value
51 – 100 0
26 – 50 1
11 – 25 2
1 – 10 3
0 4
AlterationType
Value
Cutover 0
Agriculture 1
Built Up 2
None 3
Relative Relief
Value
0 – 164 0
165 – 822 1
822 2
Topographic Variety
Value
Uniform 0
2 general topographic
types
1
> 2 topographic
types
2
Vegetative Variety
Value
1 type of vegetation
0
2 types of vegetation
1
> 2 types of vegetation
2
Water Type Value
River 0
Lake 1
Ocean 2
2
Landscape 1 Landscape 2
Flat, primarily agricultural land
(40%) with a small lake. Some
scattered balsam fir stands.
Coastal plain leading inland to
mountain plateau. Minimal
disturbance. Balsam fir and
white birch dominate
13
Model Description
Model includes 2 Python scripts in an ArcToolbox– Hex Generator– Visual Quality Model
Validation Overview
Validation
VQI Model
DataCollectionSurvey
St. John River Valley,New Brunswick
Humber River Basin,Newfoundland
Current ValidationFuture Validation
Validation
Field photographs manually rated using landscape components and associated sub-indices from visual quality model Percent and
Type of Water
Percent and Type of
Alteration
Relative Relief
Topographic
Variety
Vegetative Variety
Validation
Photographs are grouped according to ratings and/or landscape characteristics for public survey
When is a cutover no longer a cutover?
What influence
does development
have?
Validation
Public Survey– Mail routes used to define
survey zones– Intercept at Gros Morne
National Park– ~300 returns
0
2
4
6
8
10
12
14
16
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32
ModelSurvey Mean
Survey Mode
Survey Findings – Model Adjustments
Cutovers not seen as negative once regenerating (~5 years) – time since harvest needs to be included in modelStand remnants after harvest lessen negative impact of cutover – not always captured in forest inventory proceduresGreen space within built-up areas have positive impactsOpen field agriculture not as negative an impact as in literature – farms in the study area are integrated in the landscapeGolf courses not considered developedFamiliarity has slight positive impact on perceived values
Model Utility
Time 0 Visual QualityTime 0 Visual Quality
Time 30 Visual Quality – Natural Succession ScenarioTime 30 Visual Quality – Natural Succession Scenario
Time 30 Visual Quality – Forest Management ScenarioTime 30 Visual Quality – Forest Management Scenario
Integrated with other model
outputs in trade-off analysis
Model Utility
Local Scale PlanningE
xplore Options
Value change of
45%
Proposed Agriculture 15% of hex
TCH Viewshed
Next Steps
Adapt model according to survey findings.Finalize ArcGIS toolbox with updated scripts and help files.
Acknowledgements
Thank You!