Application of Logic and Decision Application of Logic and Decision - - PowerPoint PPT Presentation

Application of Logic and Decision Application of Logic and Decision Models in Models in Sustainable Ecosystem Sustainable Ecosystem Management Management

Mark Jensen, USDA Forest Service, Northern Region Mark Jensen, USDA Forest Service, Northern Region Keith Reynolds, USDA Forest Service, PNW Research Station Keith Reynolds, USDA Forest Service, PNW Research Station Keith Reynolds, USDA Forest Service, PNW Research Station Keith Reynolds, USDA Forest Service, PNW Research Station Ute Langer, University of Montana Ute Langer, University of Montana Melissa Hart, University of Montana Melissa Hart, University of Montana (HICSS (HICSS – 42) 42)

Outline for Presentation Outline for Presentation

 Describe EMDS Assessment Models  Present some R1 EMDS Assessment Models Models  Describe EMDS Decision Models  Display some R1 EMDS Decision Models

EMDS EMDS

 Ecosystem Ecosystem  Management Management  Decision Decision  Decision Decision  Support Support  System System

Background Background

 Developed by PNW Station Developed by PNW Station  Corporate USDA Corporate USDA – FS Software FS Software  Maintained by U. of Redlands and Others Maintained by U. of Redlands and Others  Maintained by U. of Redlands and Others Maintained by U. of Redlands and Others  www.Institute.redlands.edu/emds www.Institute.redlands.edu/emds  COOL STUFF! COOL STUFF!

Features Features

 Build knowledge bases for assessment with Build knowledge bases for assessment with Net Weaver Net Weaver  Build decision models for landscape Build decision models for landscape restoration restoration (treatment) with Criterion Decision Plus (treatment) with Criterion Decision Plus (treatment) with Criterion Decision Plus (treatment) with Criterion Decision Plus  Operates through GIS (ARCMAP Tool) Operates through GIS (ARCMAP Tool)  Facilitates Facilitates Transparency Transparency and and Reproducibility Reproducibility in the evaluation of monitoring data and the in the evaluation of monitoring data and the identification of treatment areas ! identification of treatment areas !

EMDS Knowledge Bases EMDS Knowledge Bases

 Evaluate Multiple Propositions that are Evaluate Multiple Propositions that are Hierarchically Designed (i.e., from the top Hierarchically Designed (i.e., from the top down where they end with data elements) down where they end with data elements) A Proposition is “Something that can be A Proposition is “Something that can be believed or denied based on the degree to believed or denied based on the degree to believed or denied based on the degree to believed or denied based on the degree to which it is true or which it is true or false” given all false” given all antecedent network conditions antecedent network conditions  Uses “ Fuzzy Membership Functions” to Uses “ Fuzzy Membership Functions” to Interpret Data Interpret Data

How to interpret map symbology How to interpret map symbology

 EMDS is an extension to ArcMap EMDS is an extension to ArcMap (ArcGIS), and its basic products are (ArcGIS), and its basic products are maps.. maps..  Each map displays the strength of Each map displays the strength of

• 1

 Each map displays the strength of Each map displays the strength of evidence (or level of support, as in evidence (or level of support, as in the legend to the left) for a specific the legend to the left) for a specific proposition. proposition.

+1

YOU Are Large

Proposition

Weight Height

Evaluation of Proposition Evaluation of Proposition

Height Weight

True, if > 6ft. True, if >200lbs True, if > 6ft. True, if >200lbs False, if < 5ft. False, if < 100lbs

Fuzzy Single Ramp Function Fuzzy Single Ramp Function

(for evaluating the proposition that watershed road density is low) (for evaluating the proposition that watershed road density is low)

Watershed Road Density

10

F (-1)

(mi/mi2) Cumulative Frequency Distribution

1 25% 50% 75% 100%

T (+1)

(-1)

Kalispell Coeur d'Alene

Strength of Evidence for Proposition

No Support [-1] Very Low Support (-1, -0.5] Low Support (-0.5, 0) Undetermined

Proposition: Road Density and # Stream Crossings are Low

Butte Missoula

Cities Undetermined [0] Moderate Support (0, 0.5] Strong Support (0.5, 1) Full Support [1] State Boundary

8 0 8 0 4 0 K il o m e t e r s

Unit: subwatershed

Fuzzy Double Ramp Function Fuzzy Double Ramp Function

(for evaluating proposition that old growth composition is within desired (for evaluating proposition that old growth composition is within desired range of conditions) range of conditions)

+1

T T

10 20 40 50

% Veg Composition

• 1

F F

Kalispell Coeur d'Alene

Strength of Evidence for Proposition

No Support [-1] Very Low Support (-1, -0.5] Low Support (-0.5, 0) Undetermined

Proposition: Composition of Old-growth Forests on Lower Subalpine Biophysical Settings is Within Desired Range of Conditions

Butte Missoula

Cities Undetermined [0] Moderate Support (0, 0.5] Strong Support (0.5, 1) Full Support [1] State Boundary

8 0 8 0 4 0 K il o m e t e r s

Unit: ecological subsection

• Some Watershed Integrity Proposition

Some Watershed Integrity Proposition Examples Examples

Watershed & Aquatic Integrity (subwatersheds) Watershed Condition Aquatic Condition Aquatic Species Threats

Watershed Condition

Impaired Water Quality (AND) Roads (mi/mi2) Components Impaired (%)

Total Impaired (%)

Stream Habitat Vegetation Upland Openings (%) Riparian Openings (%)

Upland Density

(%) (%) (%)

Habitat ( ) Sediment ( ) Temperature ( ) Metals ( ) Other ( ) Riparian Density ( ) ( ) Stream Crossings (#/mile) Low Tree Cover ( ) Seedling + Sapling ( ) Low Tree Cover ( ) Seedling + Sapling ( )

AQUATIC INTEGRITY is High

FISH SPP DIVERSITY is High

AQUATIC THREATS are Low

WATERSHED CONDITION is Good

Low Road Density and Stream Crossings

R1 Knowledge Base Design for the Evaluation R1 Knowledge Base Design for the Evaluation

• f Desired Conditions in Forest Planning
• f Desired Conditions in Forest Planning

Ecosystem Sustainability Ecosystem Sustainability

Ecological Socioeconomic Ecosystem Diversity Species Diversity Economic Social Threats Diversity Diversity

NetWeaver model NetWeaver model (cartographic (cartographic base) base) Primary propositions Primary propositions # Sub # Sub-

• proposition

proposition s evaluated s evaluated # Data # Data links links evaluated evaluated Aquatic integrity Aquatic integrity (subwatershed) (subwatershed) Subwatershed condition is good Subwatershed condition is good 19 19 13 13 Fish species status is strong Fish species status is strong 8 8 8 8 Threats to subwatershed are low Threats to subwatershed are low 17 17 12 12 Terrestrial Terrestrial integrity integrity (ecological (ecological subsection) subsection) Vegetation ecosystem diversity is high Vegetation ecosystem diversity is high 156 156 76 76 Wildlife species habitat diversity is high Wildlife species habitat diversity is high 9 9 9 9 Threats to ecological subsection are low Threats to ecological subsection are low 19 19 13 13 Fire danger Fire danger (subwatershed) (subwatershed) Fire hazard is low Fire hazard is low 8 8 5 5 Fire behavior is acceptable Fire behavior is acceptable 5 5 4 4 Fire regime is acceptable Fire regime is acceptable 4 4 3 3

Overview of NetWeaver logic model designs used to evaluate

Ignition risk is low Ignition risk is low 8 8 5 5 Social opportunity Social opportunity spectrum spectrum (subwatershed) (subwatershed) Potential for commercial uses is high Potential for commercial uses is high 6 6 5 5 Diversity of commercial uses is high Diversity of commercial uses is high 6 6 5 5 Primitive recreation opportunity is high Primitive recreation opportunity is high 5 5 5 5 Developed recreation opportunity is high Developed recreation opportunity is high 5 5 5 5 Recreation opportunity diversity is high Recreation opportunity diversity is high 5 5 5 5 Special area diversity is high Special area diversity is high 10 10 9 9 Infrastructure capacity is high Infrastructure capacity is high 24 24 17 17 Economic Economic integrity integrity (National Forest) (National Forest) Economic opportunity is high Economic opportunity is high 7 7 6 6 Overall jobs and income are high Overall jobs and income are high 6 6 4 4 Component jobs and income are high Component jobs and income are high 22 22 14 14

used to evaluate ecosystem sustainability.

Selected results of NetWeaver evaluations

Evidence that: A) subwatershed condition is good; B) road density and number of stream crossings within a subwatershed are low; C) composition of old-growth forests on lower subalpine forests on lower subalpine biophysical settings within an ecological subsection is within desired range of conditions; D) an ecological subsection has a high proportion of goshawk habitat; E) a subwatershed has high potential for developed recreation use; and F) a National Forest has high economic opportunities.

Use of Decision Models in Integrated Use of Decision Models in Integrated Landscape Protection and Landscape Protection and Restoration Restoration

• Brief Overview of R1 Integrated Restoration

Brief Overview of R1 Integrated Restoration and Protection Strategy Objectives and Protection Strategy Objectives and Protection Strategy Objectives and Protection Strategy Objectives

• Example of How EMDS Evaluation and

Example of How EMDS Evaluation and Decision Models can be Applied to this Decision Models can be Applied to this Effort Effort

Integrated Restoration and Integrated Restoration and Protection Strategy Protection Strategy Protection Strategy Protection Strategy

Northern Region’s Strategy to Protect and Northern Region’s Strategy to Protect and Restore Fire Adapted Landscapes and Restore Fire Adapted Landscapes and Watersheds Watersheds

Our Resource Focus Our Resource Focus

• Restore

Restore and maintain high value

and maintain high value watersheds

watersheds

• Restore

Restore and maintain

and maintain wildlife

wildlife habitats

habitats

• Protect

Protect communities communities and developments

and developments

Our Management Focus Our Management Focus

• Effective

Effective integration integration at various levels, within at various levels, within the agency and with our neighbors the agency and with our neighbors

• Transparent, effective

Transparent, effective priority setting priority setting process given information at multiple scales process given information at multiple scales process given information at multiple scales process given information at multiple scales

• Leading to being more effective at answering

Leading to being more effective at answering the questions: the questions: Why here? Why now? Why here? Why now?

Highlighted Risk Agents Highlighted Risk Agents

• Drought

Drought

• Bark beetles

Bark beetles

• Invasive plant and animal species

Invasive plant and animal species

• Forest encroachment into grasslands

Forest encroachment into grasslands

• Erosion and sedimentation, and toxic chemicals

Erosion and sedimentation, and toxic chemicals

• Uncharacteristically dense vegetation that creates

Uncharacteristically dense vegetation that creates hazardous fuel conditions susceptible to large wildfire hazardous fuel conditions susceptible to large wildfire

• Climate change will increase disturbance

Climate change will increase disturbance

Highlighted Resource Values At Risk Highlighted Resource Values At Risk

• Community infra

Community infra-structure structure

• Municipal watersheds

Municipal watersheds

• Watersheds and fish habitat

Watersheds and fish habitat

• Wildlife habitat, including resilient vegetation conditions,

Wildlife habitat, including resilient vegetation conditions, especially big game winter range especially big game winter range

• Use of EMDS Decision Models in the

Use of EMDS Decision Models in the Identification of Priority Areas for Integrated Identification of Priority Areas for Integrated Landscape Restoration Landscape Restoration

Decision Model Decision Model

 May use Net Weaver Assessment Results May use Net Weaver Assessment Results and other information (e.g. risks, and other information (e.g. risks, feasibility, etc…) feasibility, etc…)  Decision maker selects variables of interest Decision maker selects variables of interest  Decision maker assigns weights to Decision maker assigns weights to variables (as appropriate) variables (as appropriate) variables (as appropriate) variables (as appropriate)  May develop multiple decision models May develop multiple decision models (scenarios) to reflect different resource (scenarios) to reflect different resource concerns and public interests concerns and public interests  Output displays most efficient areas for Output displays most efficient areas for treatment and criteria for their selection treatment and criteria for their selection

Steps For Building a Decision Model Steps For Building a Decision Model

1. 1. Identify the integration unit to be used in map Identify the integration unit to be used in map display (e.g., sub watersheds) display (e.g., sub watersheds) 2. 2. Identify the scenario to be evaluated (e.g., Identify the scenario to be evaluated (e.g., watershed condition improvement, fuel watershed condition improvement, fuel reduction, etc) reduction, etc) reduction, etc) reduction, etc) 3. 3. Identify the information to be used in scenario Identify the information to be used in scenario evaluation (e.g., values, risks, and feasibility) evaluation (e.g., values, risks, and feasibility) 4. 4. Assign weights to the information used to reflect Assign weights to the information used to reflect the objectives of the scenario the objectives of the scenario 5. 5. Generate map of high priority treatment areas Generate map of high priority treatment areas

Scenario Examples Scenario Examples Scenario Examples Scenario Examples

• A. Improve watershed condition

0.40 Value 0.40 Risk

0.20 Feasibility 0.20 Watershed Integrity 0.20 Sediment 0.20 Crown Fire Potential 0.20 Roads 0.10 Forest Service Lands 0.10 Regulated Timber Base

• B. Identify fuel treatment priorities

0.40 Value 0.40 Risk

0.20 Crown Fire Potential 0.20 Beetle Damaged Trees 0.40 Wildland Urban Interface

Decision models for selecting subwatershed

• C. Protect developed recreation values

0.10 Regulated Timber Base

0.40 Value 0.40 Risk

0.20 Feasibility 0.20 # Developed Rec. Sites 0.10 Population Within 50 km 0.20 Crown Fire Potential 0.20 Beetle Damaged Trees 0.10 Forest Service Lands 0.10 Developed Rec. Opportunity 0.20 Feasibility 0.10 Forest Service Lands 0.10 Regulated Timber Base

priority treatment areas.

Kalispell Coeur d'Alene

Sub-Watershed Priority

Rules Violated Very Low Low Moderate

CDP Scenario Results: Priorities for Improving Watershed Condition

Butte Missoula

Moderate High Very High State Boundary Cities

8 0 8 0 4 0 K il o m e t e r s

Watershed ID SMART Priority Scores Criteria Legend Score

0.00 0.94 Decision Score

Highest priority subwatersheds for improvement of watershed condition.

Top 7 Priority Watersheds

• A. Improve watershed condition

0.40 Value 0.40 Risk

0.20 Feasibility 0.20 Watershed Integrity 0.20 Sediment 0.20 Crown Fire Potential 0.20 Roads 0.10 Forest Service Lands 0.10 Regulated Timber Base

• B. Identify fuel treatment priorities

0.40 Value 0.40 Risk

0.20 Crown Fire Potential 0.20 Beetle Damaged Trees 0.40 Wildland Urban Interface

Decision models for selecting subwatershed

• C. Protect developed recreation values

0.10 Regulated Timber Base

0.40 Value 0.40 Risk

0.20 Feasibility 0.20 # Developed Rec. Sites 0.10 Population Within 50 km 0.20 Crown Fire Potential 0.20 Beetle Damaged Trees 0.10 Forest Service Lands 0.10 Developed Rec. Opportunity 0.20 Feasibility 0.10 Forest Service Lands 0.10 Regulated Timber Base

priority treatment areas.

A) improving watershed condition, B) hazardous fuel reduction in the wildland-urban interface, C) minimizing fire and

CDP Scenario Results – Landscape Restoration

Suggested priorities for: C) minimizing fire and bark-beetle hazards within areas with high developed recreation

• pportunities, and

D) composite priority scores averaged across all three decision models (A + B + C).

Other potential scenarios Other potential scenarios

• Improvement of watershed condition

Improvement of watershed condition

• Improvement of bull trout habitat

Improvement of bull trout habitat

• Protection of USFS infrastructure

Protection of USFS infrastructure investments in fire prone areas investments in fire prone areas investments in fire prone areas investments in fire prone areas

• Protection of private homes and

Protection of private homes and infrastructure in WUI infrastructure in WUI

• Improve species habitat for species of

Improve species habitat for species of concern and interest concern and interest

• Others?

Others?

*

Integrated Restoration & Protection Priority (subwatersheds) Wildlife Community Fire Resilience Recreation Aquatic Species Watershed Condition Vegetation ?

Developed Recreation

*

Scenic Integrity

* *

Dispersed Recreation Fish

* *

Water Quality (Sediment)

Communities

• f Concern

*

Special Areas

* *

Rare Plants

Unique Habitats

*

Themes & Subthemes (Goals) Objectives (Scenarios)

KEY

Discussion Time Discussion Time