alaska’s changing fire regime –implications for the vulnerability of its boreal forests
DESCRIPTION
Alaska’s Changing Fire Regime –Implications for the Vulnerability of its Boreal Forests. BC LTER Annual Symposium 26 February 2010 E. Kasischke, D.Verbyla, S. Rupp, D. McGuire, K. Murphy, R. Jandt, J. Allen, E. Hoy, P. Duffy, M. Calef, M.Turetsky. Outline. - PowerPoint PPT PresentationTRANSCRIPT
1
Alaska’s Changing Fire Regime –Implications for the Vulnerability of its
Boreal ForestsBC LTER Annual Symposium
26 February 2010
E. Kasischke, D.Verbyla, S. Rupp, D. McGuire, K. Murphy, R. Jandt, J. Allen,
E. Hoy, P. Duffy,M. Calef, M.Turetsky
2
Outline
1. Sources of fire regime information
2. Variations in the fire regime over time
3. Satellite observations of fire4. Fire and vulnerability of black
spruce forests
3
Measurement of fire regime characteristics
1. Fire ignitions – lightning strikes, fire management records
2. Burned area – Annual fire reports, fire event databases, fire perimeters, satellite observations, theoretical backcast models
3. Fire frequency – burned area, tree rings, sediment charcoal (not addressed in paper)
4. Seasonal patterns of burning – daily fire management reports, satellite observations, fire event reconstructions
5. Fire severity – field observations, satellite observations
4
1. Beginning in the late 1940s, the availability of surplus aircraft plus trained pilots increased access to remote fire events
2. High-quality base maps (USGS) were not available until the 1970s
3. Since the early 1990s, the availability of GPS has improved mapping of fire perimeters considerably
4. In some instances, Landsat data are being used to map fire perimeters
Kasischke et al. IJWF, 2002
Quality of burned area data is highly variable over time
5
Old data is not necessarily bad data
6
0
1
2
3
1860 1890 1920 1950 1980 2010
Bu
rned
are
a (m
illi
on
ha)
Large-fire year threshold
7
0
200
400
600
800
1,000
Bu
rne
d a
rea
(1
00
0 h
a/y
r/d
ec
ad
e)
0
1
2
3
4
5
La
rge
fir
e y
ea
rs p
er
de
ca
deBurned area
Large fire years
8
0
0.1
0.2
0.3
0.4
0.5
0.6
May June July August September
Fra
ctio
n o
f li
gh
tnin
g s
trik
es
All years (66 k/yr)
Large fire years (78 k/yr)
Small fire years (61 k/yr)
0
100
200
300
400
500
1960s 1970s 1980s 1990s 2000s
An
nu
al n
um
ber
of
fire
s
Lightning ignitedHuman ignited
Lightning strike data available for 1990s, 2000s
Sensor technology changed in late 1990s
Human vs. lightning ignitions
9
0
100
200
300
400
500
600
1950s 1960s 1970s 1980s 1990s 2000s
La
rge
st
fire
ev
en
t (x
10
00
ha
)
0
5
10
15
20
25
30
35
40
45
50
Nu
mb
er
of
fire
ev
en
ts >
50
,00
0 h
a
Human event
Lightning event
Human events
Lightning events
Studies by DeWilde and Chapin 2006 and Calef et al. 2008 show
As proximity to human infrastructure/access (settlements, roads, rivers) decreasesThe number of human ignitions increase Burned area decreases
10
0
200
400
600
800
1940s 1950s 1960s 1970s 1980s 1990s 2000s
Bu
rned
are
a (1
000
ha
/ yr
)
TotalHuman-caused firesLate-season fires
4 3
2
1
1
3
4
11
PSD*
*Primary sampling device
Our research is based on data collected in plots in 36 fire events and unburned stands • Used to measure fire severity and surface fuel consumption in black spruce forests
Studies by researchers at ERIM, UMD, MSU, UAF, USFS, USGS, USFWS
284 plots in unburned stands, 465 plots in burned stands8,447 organic layer depth measurements in unburned stands, 10,140 in burned stands
>2,000 organic layer samples collected for lab analysis to determine bulk density and % C
12
Alaska Black Spruce Forests
0
10
20
30
Backslopes Flat uplands Flat Lowlands Weightedlandscape
Dep
th o
f S
OL
bu
rnin
g (
cm) Early season
Late Season
Turetsky et al. in review
13
Geospatial data sets for analysis of patterns of burning
14
y = 0.78x
R2 = 0.96, p < 0.0001
y = 0.32x
R2 = 0.93, p < 0.0001
0
10
20
30
40
50
0 10 20 30 40 50Burned + unburned area (x1000 ha)
Bu
rned
are
a (x
1000
ha)
2004 fires2006 fires
2004 Fires
y = 0.78x
R2 = 0.97, p < 0.0001
0
50
100
150
200
250
0 50 100 150 200 250
Burned + unbured area (x 1000 ha)
Bu
rned
are
a (x
100
0 h
a)
15
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Spruce forest Deciduousforest
Mixed forest Non-forest
Fra
cti
on
of
lan
ds
ca
pe
Landscape
Burn Perimeter
Burned area
Landscape = Interior forest ecoregionsBurn = 2004 fires
16
1900 20001980
Black spruce forests that have burned over the past 25 years were formed ~80
to 200 years ago under much different climatic
conditions
Significant changes in the fire regime have occurred over
the past 25 years
1. To what extent will black spruce forests reorganize themselves in response to the changes in the
fire regime and climate?
2. What processes will control this re-organization?
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1940 1950 1960 1970 1980 1990 2000 2010
An
nu
al b
urn
ed a
rea
(mil
lio
n h
a)
0.0
0.2
0.4
0.6
0.8
1.0
10-y
ear
aver
age
bu
rned
are
a (m
illi
on
ha)
decrease of 5,400 ha/yr
increase of 23,600 ha/yr
17
Controls on Vulnerability of BS Forests
1. Topography – lowland sites are resilient to deep burning fires
2. Frequency of large fire years a. A higher fraction of the landscape burnsb. More large fire events increased
depth of burning in flat uplands and backslopes in early and mid-season fires
c. More late season burning increased depth of burning in flat uplands and backslopes