Landscape - Level Fire Characteristics

advertisement
Landscape-Level Fire
Characteristics
in Spanish Creek Watershed
UC Berkeley
Fire Science Lab
Professor Scott Stephens
Dr. Brandon Collins
Jason Moghaddas
Dr. Emily Moghaddas
Dr. Kurt Menning
Research Focus
How do fuel treatments affect landscape-level potential
fire behavior in the Spanish Creek Watershed?
2007 Wheeler fire from
Genesee Valley
General Study Area
Spanish Ck watershed,
Treatments completed
2008 Rich
Fire
Mixed conifer forest
**
Fuel treatments:
Meadow Valley Project
Several wildfires in past
decade: 2008 Rich Fire
Assisted in calibration
of fire modeling
**
**
Study Area Detail
46,000 acres
Multiple Land Allocations:
Deferred -1%
Offbase <1%
Spotted Owl Habitat Area -1%
Home Range Core Area -14%
Protected Activity Center - 9%
Riparian Habitat Conservation Area
(150’ & 300’ buffers) -12%
Defensible Fuel Profile Zone - 9%
Group Selection - 1%
Other National Forest Lands – 53%
Building the “Virtual Landscape”
High resolution IKONOS imagery used to assess fine grain forest
structure and build base FLAMMAP and FARSITE landscape
Overlaying actual treated DFPZ’s and Groups in the Meadow Valley
Project
FLAMMAP values were adjusted to reflect the range of
values observed in pre & post-treatment field data collected from the
Guard and Waters projects
FLAMMAP landscape further
calibrated to accurately “remodel”
the approximate the size and postfire effects of the 2008 Rich fire
This was important
Burning with FLAMMAP
The landscape was burned at 97th percentile weather
conditions
Weather Calculated from Cashman RAWS, filtering data
for all weather between, June 1 - Sept 30, 2002-2008
Windspeed from
Cashman RAWS
Fuel Moisture (%)
1-hr
10-hr
100-hr
1,000-hr
Herb
Woody
1-Minute
Windspeed
Wind
direction
1.2%
2.1%
5.5%
6.4%
35.4%
60.7%
25 mph
2250
Winds vectored to 10m accuracy using “WindNinja”
Conditional Burn
Probability Parameters
Fire probability in study area: FLAMMAP
30 meter node resolution
Max simulation time three 5-hour burning periods
(900 minutes)
Number of random ignitions = 1,000
Modeling results: Surface Fire
% Area modeled as surface fire
100
75
50
25
151 ac
302 ac
1917 ac
3097 ac
SOHA
Deferred
PAC
HRCA
1822 ac
336 ac
10 ac
4059 ac
11706 ac
34 ac
DFPZ
Other NFS
lands
Offbase
0
Perennial
Intermit,
Group
streamzone ephem zone Selection
Land Allocation
% Area modeled as passive crown fire
Modeling results: Passive Crown Fire
100
75
50
25
21 ac
38 ac
845 ac
1255 ac
PAC
HRCA
1556 ac
209 ac
559 ac
8 ac
5493 ac
45 ac
Other NFS
lands
Offbase
0
SOHA
Deferred
Perennial Intermit,
Group
streamzone ephem zone Selection
Land Allocation
DFPZ
% Area modeled as active crown fire
Modeling results: Active Crown Fire
100
75
50
25
296 ac
237 ac
1284 ac
2022 ac
SOHA
Deferred
PAC
HRCA
584 ac
123 ac
3 ac
8 ac
4812 ac
26 ac
Other NFS
lands
Offbase
0
Perennial
Intermit,
Group
streamzone ephem zone Selection
Land Allocation
DFPZ
Change in Burn Probability
Post treatment – Pre
treatment landscape
Green color, lower
probability of burning
Treated areas have an
impact on area
Burn Probability by land allocation
Pre-treatment
Probability
0.3
Post-treatment
0.2
0.1
481 ac
595 ac
4110 ac
6522 ac
SOHA
Deferred
PAC
HRCA
4538 ac
706 ac
568 ac
4087 ac
24299 ac
DFPZ
Other NFS
lands
117 ac
0.0
Perennial
Intermit,
Group
streamzone ephem zone Selection
Land Allocation
Offbase
Meadow Valley FARSITE fire simulation
(fire coming up from Middle Fork Feather River Canyon)
Initial Findings
All land allocations had majority of fire type modeled as passive
or active crown fire before treatment
No clear trends among untreated land allocations - all show
similar results at the landscape scale
46% of deferred modeled as having passive and active crown fire
DFPZ’s were modeled as surface fire
Group selections were modeled as passive crown fire
Wildfire from middle fork of Feather River moderated by fuel
treatments in Meadow Valley
Only 10% of area treated, low end of recommendations
Manuscript in development on project
Fuel Treatment Longevity in
the Northern Sierras
Lindsay A. Chiono
Ph.D. Candidate
Objectives
• Characterize vegetation and fuels development over time
based on a chronosequence of fuel treatments.
• Sample across fuels reduction methods, forest types, and
site qualities to assess their respective influences on treatment
longevity.
• Develop projections for stand and fuel development for a
range of fuel treatments.
• Develop treatment regimes for establishment and
maintenance of DFPZ’s.
0 to 5 years
Thin Only, North Aspect
Thin Only, South Aspect
Thin/Burn, North Aspect
Thin/Burn, South Aspect
6-10 years
11+ years
50 Treatments Sampled:
Almanor RD: 2
Mt. Hough RD: 14
Beckwourth RD: 3
Sierraville RD: 6
Truckee RD: 16
Collins Pine: 9
Preliminary Data

Data from 16 treatments sampled in 2007 on the
Plumas National Forest were analyzed to
illustrate the information collected and the
modeling of potential fire behavior.
Site Characteristics: Fuel Loading
Fuel Loading
5
tons/acre
4
100 hr
3
10 hr
2
1 hr
1
0
3
4
5
5
10
10
11
11
13
24
25
26
Age (yrs)
1000-hr Fuel Loading
60
tons/acre
50
40
30
20
10
0
3
4
4
5
5
10
10
Age (yrs)
11
13
24
25
26
% Cover
Shrub Cover
80
70
60
50
40
30
20
10
0
North Aspect
South Aspect
0
5
10
15
20
25
30
Age (yrs)
% Canopy Cover
Canopy Cover
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
0
5
10
15
Age (yrs)
20
25
30
Analysis in FMAPlus

FMAPlus allows the estimate of potential fire behavior and effects based on
fuels and weather data.

Inputs: tree measurements, fuel model, 90th and 97th percentile weather
calculated from 45 years of historical RAWS data.

Fire Behavior Outputs: include rate of spread, flame length, torching and
crowning indices, and probabilities of tree mortality.
Photos courtesy of the Stephens Lab
Flame Length
Flame Length (ft.)
6
Modeled flame
length and fire rate of
spread similar
regardless of age of
treatment
5
4
3
2
1
0
0
5
10
15
20
25
30
Age (yrs.)
Similar results for
Torching and
Crowning indices
Rate of Spread
60
ROS (ch/hr)
50
40
30
20
10
0
0
5
10
15
Age (yrs)
20
25
30
Summary

Fuel breaks may be fairly long-lived in these forest
types (full analysis ongoing)

Stratification by site quality, forest type, and
treatment type may help clarify trends

Results applicable to mixed conifer forests near
Quincy and to the east, not west-side mixed conifer
forests that are more productive

Analysis of data continues
Acknowledgments
Longevity Project: Thanks to Matt Cerney, Scott Conway, Mike De
Lasaux, Jay Francis, Jon Lamb, and Bruce Troedson for guidance and
help in locating study sites, and to field assistants Jordan Aney, Anton
Chiono, Marisa Parish, and Rebecca Susko.
Funding provided by the BLM through the California Fire Safe Council
Grants Clearing House, the UC Division of Agriculture and Natural
Resources (DANR), and the Sierra Nevada Conservancy.
Landscape Project: Thanks to PNF Employees Colin Dillingham, John
Yembu, Ryan Tompkins, Steve Causemaker, Gary Rotta, and Ryan
Bower for providing treatment unit, monitoring, and owl data. Funding
Lassen-Plumas Study (USFS)
stephens@nature.berkeley.edu
http://www.cnr.berkeley.edu/stephens-lab/
Download