Integrating Climate Change into Landscape Planning

Modeling climate and management interactions within the ILAP framework

April 23, 2013Jessica Halofsky

Emilie Henderson

Modeling

• “Essentially, all models are wrong, but

• some are useful.” – Box & Draper 1987

“It’s Only a Model.” – Patsy, 1975

Projects behind today’s talks

• Integrated Landscape Assessment Project (ILAP)– Climate Change Module– Central Oregon Study Area

• Climate, Management and Habitat

How might climate and land management interact to shape vegetation and habitat?

Coastal Washington

Southwest Oregon Southeast Oregon

Northern Spotted Owl Greater Sage Grouse

Scenarios

General Topics for today’s talk:

• Starting conditions • STMs without climate• Climate impacts modeling

Modeling Strata

The overall picture:

Not locally precise.Useful for describing landscape-to regional

trends.

Current Vegetation

• GNN = Gradient Nearest Neighbor

– A spatial depiction of the FIA plots. – Structured by an ordination model.

• Gives us information on current vegetation within each modeling stratum.

Janet Ohmann, Matthew Gregory, Heather Roberts

General Topics

• Starting Conditions • State Transition Modeling, without

accounting for climate• Climate impacts modeling

State and Transition Modeling

Early Successional Young Forest

Mature Forest

Old Growth Forest

GrowthFireRegeneration Harvest

State and Transition Modeling:Dry Douglas-fir

Grass-Forb

Giant TreesModerate CanopyMulti-Layered

Pole-stage, single-story, post-disturbance

ILAP Potential Vegetation Types

PROBLEM:

Basic framework assumes that this map doesn’t change.

When climate shifts, sowill potential vegetation.

State and Transition Modeling

Early Successional Young Forest

Mature Forest

Late Successional Old Growth

GrowthFireRegeneration Harvest

Estimated harvests from the LANDSAT record inSouthwest Oregon

19

84

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

year

0

20000

40000

60000

80000

100000

120000FederalPrivate Industrial

Acr

es

Ha

rve

ste

d

Thanks to Robert Kennedy for the LandTrendr maps of disturbance history

Current Fire

19

84

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

year

0

100000

200000

300000

400000

500000

Acr

es

bu

rne

d

PROBLEM:

Fire regimes are set to resemble the recent past.

They will probably change with shifting climate.

Extracted from Monitoring Trends in Burn Severity dataset: mtbs.gov

Preliminary results for Southwestern Oregon: Current management

Area with Large and Giant TreesNo Climate Change

General Topics

• Starting Conditions• State Transition Modeling, without accounting

for climate• Climate impacts modeling

– ILAP extension – Central Oregon Study Area

What about climate change?

• Climate controls ecosystem processes, including:

– Plant establishment, growth, and mortality

– Disturbance• Drought• Fire • Insect outbreaks

Dynamic Global Vegetation Models (DGVMs):

• Link state-of-the-art knowledge of: – plant physiology– biogeography – biogeochemistry– biophysics

• Simulate changes in vegetation structure and composition and ecosystem function through time

*adapted from: Bachelet, D., J. M. Lenihan, C. Daly, R. P. Neilson, D. S. Ojima, and W. J. Parton. 2001. MC1: A Dynamic Vegetation Model for Estimating the Distribution of Vegetation and Associated Ecosystem Fluxes of Carbon, Nutrients, and Water. USDA Forest Service General Technical Report PNW-GTR-508.

biomass mortality

nutrient loss and release

Biogeography Biogeochemistry

Fire

fireoccurrence

lifeformmixture

carbon poolssoil moisture

lifeform mixture

live biomass

(MAPSS) (CENTURY)

(MCFire)

The MC1 Dynamic Global Vegetation Model

A Linked Model Approach

Dry Mixed Conifer

Xeric Ponderosa PineJuniper woodland

Moist Mixed Conifer

MC1STMs

Central Oregon Study Area

Historical vegetation in the study area

Vegetation type crosswalks

MC1 Functional Vegetation Type

STM Potential Vegetation Type

Subalpine forest Mountain hemlock and subalpine fir forests

Cool needle-leaved forest Moist mixed conifer and white fir forests

Temperate needle-leaved forest Ponderosa pine, lodgepole pine, and dry mixed conifer forests

Temperate needle-leaved woodlandMountain big sage – western juniper woodland and shrubland

Temperate shrubland Wyoming big sage shrubland

Xeromorphic shrubland Salt desert shrubland

Temperate grasslandBluebunch wheatgrass – Sandberg bluegrass grassland

Warm-season grassland Warm-season grassland

Climate Scenarios

MC1 Functional Vegetation Type Projections

MIROC CSIRO

Hadley

Halofsky et al. in review

MC1 fire projectionsMIROC CSIRO

Hadley

Halofsky et al. in prep

Linked model results

MIROC CSIRO

Hadley

Halofsky et al. in prep

Central Oregon Management Scenarios

• Fire suppression only– Fire frequencies same as the last 25 years under

fire suppression– No other active management

• Resilience– Light to moderate levels of thinning and some

prescribed fire in dry forest types

Effects of management on dry forests

Fire suppression

only

Resilience

MeanMin to maxRandomly selected simulations

Halofsky et al. in prep

Fire suppression only

Resilience

Halofsky et al. in prep

Effects of management on:dry forests with large trees and open canopy

Land

scap

e pr

opor

tion

Land

scap

e pr

opor

tion

Fire suppression

only

Resilience

Trends in dry forest structure<12.7 cm12.7-50.8 cm>50.8 cm

Conclusions for Central Oregon

• Linked DGVM-STM output suggests greater vegetation resilience than DGVM alone.

• Dry ponderosa pine and mixed conifer forests will likely maintain dominance in the central Oregon study area.

• In some cases, management may dampen the magnitude of forest change under changing climate.

Halofsky et al. in prep

Technical Thoughts

• All models are wrong, ours could be useful• These models provide big-picture projections• The linked model process is data-, labor-, and

software-intensive

Getting to Landscape Planning

• We haven’t described the planning process itself, which involves conversations and people.– Stakeholders – Decision Makers

• Our models are useful storytelling tools. – Enable the asking of questions.– Realistic and plausible stories. – Enhance the role of science in conversations about

planning.

• Half of science is asking the right questions. -- Roger Bacon (c. 1214 – 1294)1

• Emilie, you really need to refine your questions!-- Dr. David Mladenoff, numerous times throughout my

career as a PhD student in his lab.1Wikiquotes

19

84

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

20

07

year

0

20000

40000

60000

80000

100000

120000FederalPrivate Industrial

Acr

es

Ha

rve

sted

What activities?e.g., partial harvestregeneration harvestrestoration harvestprescribed fire

At what rates?

Where should they be applied?

Groups we have spoken with:

•The Nature Conservancy•Bureau of Land Management•US Forest Service personnel – regional and local•Local chapters of the Society of American Foresters•Washington Department of Natural Resources•Oregon Department of Forestry•Consulting foresters who serve nonindustrial private landowners•County commissioners

•4 activities

• 24 ownership/allocation categories

• ∞ variations in rates

Save the world!

Our Hope for our Work

• Tell informative stories that are grounded in science about how different landscape management policies and plans may lead to different futures.

• Relevance and credibility beyond the science community.

Funding:

Dominique Bachelet Emilie HendersonDavid Conklin James KaganMegan Creutzburg Becky KernsJessica Halofsky Anita MorzilloJoshua Halofsky Janine SalwasserMiles Hemstrom

ResearchTeam: