hong s. he university of missouri-columbia
DESCRIPTION
LANDIS 4.0, A New Generation Computer Simulation Model for Assessing Fuel Management Effects on Fire Risk in Eastern U.S. Forest Landscapes. Hong S. He University of Missouri-Columbia. Acknowledgements. Original Design: David Mladenoff LANDIS 4.0 Dynamic Design: Hong He - PowerPoint PPT PresentationTRANSCRIPT
LANDIS 4.0, A New Generation Computer Simulation Model for Assessing Fuel Management Effects on Fire Risk in Eastern U.S. Forest LandscapesHong S. He
University of Missouri-Columbia
Acknowledgements Original Design: David Mladenoff LANDIS 4.0 Dynamic Design: Hong He
Succession/Dispersal Module David Mladenoff, Hong He
Fire Module Jian Yang, Hong He, Eric Gustafson
Wind Module David Mladenoff, Hong He
Harvest Module Eric Gustafson, Stephen Shifley, Kevin Nimerfro, David Mladenoff,
Hong He Fuel Module
Hong He, Bo Shang, Thomas Crow, Eric Gustafson, Stephen Shifley Biological Disturbance Module
Brian Sturtevant, Eric Gustafson, Wei Li, Hong He LANDIS 4.0 Programming
Vera W. Li, Jian Yang, Bo Shang, and Hong He Funding Supports for LANDIS 4.0 Development
USFS North Central Research Station
Introduction
National/Regional Landscape Stand
Tools• Fuel Characteristic Classification (FCC)
• Fire Regime Condition Class (FRCC)
• National Fire Danger Rating System (NFDRS)
• Other tools
Results• Fire, fuel conditions• Management guidelines• Other results
Tools• Decision Support System• Landscape fire succession
Models• Other tools
ResultsManagement Plans• Short-term (10-20 yr)• Mid-term 20-50 yr)• Long-term >50 yr
Tools• BEHAVE• FVS-FFE• CONSUME• FOFEM• Other tools
Results• Fuel consumption• Smoke production• Fire effects• Other results
See reviews by Keane et al (Ecol. Model. 179) and Mladenoff (Ecol. Model. 180, 2004)
LANDISLandscape Disturbance and Succession ‘Stochastic cellular automaton’ Raster-based
Complex spatial dynamics computationally possible Infinite aggregation and dissolution of patches
Spatial Scales Functional extent 100s ha – 107 ha Functional resolution- cell size 10x10m - km2
Temporal Scales Current model time step is 10 yr, while a version of annual time
step has been developed and is being used in Southern California.
LANDIS simulates multiple disturbance and management processes in combination with the simulation of succession dynamics at the tree species level.
LANDIS 4.0 Model DynamicsTree Species Establishment
and Resprouting
Species LevelCompetitive Succession
and Dispersal
Disturbance and Harvest Related
Mortality
Wind(top down)
Fire(bottom up)
Insect/Disease(specie/age specific)
Harvest(specie/age specific)
Fuel(Accumulation/reduction)
Spec
ies
age/
size
su
scep
tibili
ty
Modify fine/coarse fuel
Modify Specieand Age Cohorts
Land type
Shade tolerance
Backgroundmortality
Longevity
Dispersal distance
Disturbance and Management
Modified from Mladenoff 2004 Ecological Modelling
LANDIS Operational Design
climate zone
soil mapDEM
Multiple fire regimes: Ignition, size, cycle, spread, intensity and severity
multiple species and age input maps
output singlespecies map
year 0
year n
reclassifiedvegetation type
output disturbances
species age Classes
model input
model simulation processes
Environmental boundariesand constrainsSite and species interactionssuccession, seeding, disturbance history, and disturbance interaction
Harvest prescriptions: stands, management units, rotationsize, species, and methods
fire
wind
harvest
Land type
Insect/disease
fuelFine, coarse and life fuelAccumulation/decomposition
Wind regimes: size, cycle, spread, intensity and severity
Epicenter, frequency, size, Hosts, susceptibility, intensity, and severity
LANDIS 4.0 Software Design--Component-based
SUCC
ESSI
ON
1/O
BDA
I/O
FIR
E
I/O F
UEL
I/O
HARV
EST
I/O
User1
User2
Usern
species name longevitymaturity
mortality growth
dispersalsprouting
fire toleranceshade tolerance
inte
rnal
des
ign
I/O
WIN
D
Sites of LANDIS Applications
Modified from Mladenoff 2004 Ecological Modelling
An Application in Missouri Central Hardwood Forests
Fuel and Fire Management Issues Over half a century fire suppression Increasing fire intensity Need for fuel management
Currently prescribed burning 0.06% per year for fuel reduction
How extensive should fuel treatment be expanded, 0.6%, 1.2%, or 2.4% per year?
Should coarse woody debris reduction be employed?
What are the effects of the combinations of fuel treatment size and method?
Simulation Design
Treatment Size (percentage /yr)
Trea
tmen
t M
etho
d
0.06 0.12 0.6 1.2 2.4 4.8
PB
PB+CWD
Factorial Design of Fuel Treatments
A B C D
E F G H
Simulation Design Using existing data of species/age and land
types Current fire regime under suppression
Mean fire size 3.5=ha, SD=1.8 ha Fire cycle 300 yrs on southwestern slopes and 450
yrs on northeastern land types Harvest
Even/uneven aged harvesting on oldest forest (>100 yrs) and “thinning” on young forest (<30 yrs), 0.4% /yr
Statistical Analysis Each treatment was simulated to 200 years
with 10 replicates Overall treatment effects were analyzed using
multivariate analysis of variance (MANOVA) Treatment pair comparisons was analyzed
using ANOVA Response variables:
percent pixels (% landscape) simulated fire severity classes (1-5), with 4 severely
stand damaging and 5 stand leveling fire five age classes of four species group, black oak,
white oak, shortleaf pine, and maple Seedling, sapling, pole, sawlog, and old-growth
Results and DiscussionSimulated Fire and Species
Dynamics
Results and DiscussionSimulated Fuel Dynamics
Simulated Fires at Different Severities
200 year means
0
10
20
30
40
50
A B C D E F G HTreatment
Fire
s B
urne
d on
the
Land
scap
e (%
)
severity 5
severity 4
severity 3
severity 2
severity 1Ground fire
Severe damaging
Stand leveling
Medium damaging
Low damaging
Simulated Fires at Different Severities
200 Year Dynamics
Simulation Years
Fire
Sev
erity
Cla
ss P
ropo
rtio
ns
(%)
020406080
10 40 70
100
130
160
190
12345
H)Fire severity
020406080 A)
020406080 B)
020406080 C
0
20
40
60
80
10 30 50 70 90 110
130
150
170
190
D)
020406080 E)
020406080 F)
020406080 G)
0
20
40
60
8010 30 50 70 90 11
0
130
150
170
190
H) Fire severity
Effects of Fuel Treatments on Fire Severity
Land
scap
e (%
)
Land
scap
e (%
)
A B C D E
F
G
H
A B C D E
F
G H
A B C DE F
GH
A B C DE
F G H
AB
CD
E
F
GH
Ground fire
Severe damaging
Stand levelingMedium damaging
Low damaging
Simulated Age Class Compositions
200 year means
0%
20%
40%
60%
80%
100%
A B C D E F G HTreatment
Age
Cla
sses
old-growthsawlogpolesaplingseedling
0%
20%
40%
60%
80%
100%
A B C D E F G HTreatment
Age
Cla
sses
old-growthsawlogpolesaplingseedling
0%
20%
40%
60%
80%
100%
A B C D E F G HTreatment
Age
Cla
sses
old-growthsawlogpolesaplingseedling
0%
20%
40%
60%
80%
100%
A B C D E F G HTreatment
Age
Cla
sses
old-growth
sawlog
pole
sapling
seedling
0%
20%
40%
60%
80%
100%
A B C D E F G H
old-growthsawlogpolesaplingseedling
Black oak Pine
Maple White oak
Effects of Fuel Treatments on Species Age Compositions—Seedling/Sapling
Black oak Pine
Maple White oak
Land
scap
e (%
)La
ndsc
ape
(%)
B CD
E
F GH
A
B CD
E
F
GH
A
B C D E F G
H
A
Effects of Fuel Treatments on Species Age Compositions—Old Growth Black oak Pine
Maple White oak
Land
scap
e (%
)La
ndsc
ape
(%)
B C D EF
GH
A
B C DE
F
GH
A
B C D E
F GH
A
BC
D
E
FG
H
A
Summaries Low and mid severity fires will change to severe
stand damaging (class 4) or stand leveling (class 5) fires under the situation of limited fuel treatment (e.g., 0.6%/yr) in central hardwood forests.
Fuel treatments reduce fire occurrence, prolong fire cycle, and reduce the high severity fires. Fire cycles extend from current 300-450 years to over
1,000 years depending upon treatment methods. Total sites burned by high severity fires are reduced from
11% to 1% of the study area. Fuel treatment size exhibits threshold effects in the
study area: When treatment size is small (e.g., <0. 6% per year
(treatment B), treatment methods have little effect in terms of reducing fire severity.
Summaries CWD reduction in combination with prescribed burning is
more effective in reducing fire severity.
The benefits of fuel treatment are not linearly related to the treatment effort. Increasing treatment size from 1.2%/yr (F) to 2.4% /yr (G)
reduces severe stand damaging fire from 2.4% to 0.5%. Further increasing treatment size to 4.8% (H) had only small effects (although statistically significant) on high severity fires. Thus planners and managers should consider a balanced approach in terms of treatment size and treatment effects.
PBPB
CWD
Treatment F Treatment DTreatment effects
Summaries Fuel treatment effects on species composition
and age class are statistically significant, except when treatment size is too small (e.g., <0.6%). However, such effects are secondary to the effects of fire suppression on species composition and age structure (result not shown).
We showed that LANDIS is a effective model for evaluating long-term and large spatial effects for the identified scenarios. The actual future is only one realization of numerous possible scenarios that are beyond any simulation capacity. Thus LANDIS is not a predictive model. Nevertheless, the model is useful since the large-scale effects would otherwise be very difficult to assess.