Yellow-Cedar Decline:
Terrain Factors Characterize Decline Distribution and Hint of Climate Influence
2
a)
Elevation
b)
Aspect
c)
Slope
Cedar-decline
North
High
High Northeast
East
Low
Southeast
Dustin Wittwer, USFS, FHP
Paul Hennon USFS, FHP/PNW
Dave D’Amore USFS, PNW
Melinda Lamb USFS, FHP
Evaluati
ng Key Landscape Features of a ClimateEvaluating
Climate-Induced Forest Decline
Low :
South
Southwest
West
Yellow-cedar, the most economically valuable tree in Alaska, has been subjected to wide
spread decline during the last century. R ecent evidence suggests this decline is climate
induced, related to a reduction of an insulating snow pack in winter and early spring. We
chose Mt. Edgecumbe, an inactive volcano near Sitka, Alaska, to test the influence of various
terrain factors. Unique radial symmetry, relatively even slope gradient, homogeneous soils and
apparent presence of yellow-cedar forests from sea level to timberline presented an ideal place to
study decline distribution.
Project Funding by Evaluation & Monitoring
CIR image drape
Project WC-EM-07-01; Year 1 of 2
Northwest
3
1000
terrain/decline relationships were a factor of recent
climate influences or simply driven by the
presence or absence of live cedar.
0.120
*area
600
500
400
300
200
100
0.050
*area
1
0.100
0.150
0.08 0
NE
0.06 0
0.04 0
0.02 0
W
E
0.00 0
More decline
occurs on the
warmer
southern
aspects.
Steeper well-
drained slopes support more
insulating
canopy &
deeper
roots.
0.080
0.060
0.040
0.020
SW
SE
S
0.200
Aspect:
calculated from
30 meter DEM
20% - 40%
40% - 60%
60% - 80%
0.000
2
6 10 14 18 22 26 30 34 38
Photo Interpretation & Aerial Survey Verification Improve
Cedar Decline Mapping Resolution
a) On some landscapes, like Mt.
Edgecumbe, where cedar-decline
occurs almost continuously.
Discrete pockets are too numerous and small to effectively map from
an aircraft moving
at 100 mph.
2a
3a
To assess the extent of live yellow
cedar at and above the limits of the
mapped decline, we set up 96 virtual
plots equally stratified in eight cardinal
wedges and along four contours from
600 ft elevation to 1800 ft (timber line).
We recorded relative species composition and
health from a helicopter, using the digital aerial
sketchmap system. Results were interpolated
into a map showing percentage of live
yellow-cedar.
g
pin
ap
rm
da
ce
h
alt
he
elev
atio
n
•Cedar
•Healthy
•1800 ft
Photo interpretation
Aerial survey
3b
ing
ns
se
ote
r em
1b
Healthy yellow-cedar forests
clearly extend beyond the
elevation limits of the decline.
grou
nd c
hec
king
n
tio
eta
rpr
n te
i
o to
ph
Aerial surveys verified &
refined the PI produced data.
The PI
method
produced
both false
positives
and false
negatives.
Although dead yellow-cedar can be
mapped by air and detected on
remotely sensed images, the
presence and abundance of live
yellow-cedar in healthy forests is
not easily distinguishable from other
species and not currently quantified
or mapped on a broad scale.
none
up to 20%
2b
*Each chart represents the acreage of cedar-decline normalized against the acreage of the terrain factor slope (deg)
b) Because aerial survey
methods yield a lot of “lumped”
data, we produced a photo
interpreted (PI) dataset for a finer
scale analysis using 1:60,000 CIR
photography. We then aerially
verified the PI data using Digital
Aerial Sketch-map techniques.
Sample Points
Live Cedar
Kriging results
2c
aer
ial s
urv
ey
0.000
NW
ct
pe
as
When the tree
roots are most
susceptible to
freezing, the
protective snow
layer no longer
persists at the
lower elevations
700
*
#
0.100
N
0 .10 0 slop
e
micro asp ect
800
*area
900
elevation (ft)
Decline Distribution is not Simply Defined by the Extent of
Yellow-Cedar Forests
a) We needed to determine whether or not the observed
b) Ground plot transects verified
our helicopter assessments of
forest composition.
Ground evaluation verified
healthy yellow-cedar on all
aspects and a range of slopes.
4 Remote Sensing Methodology Will Utilize UAVs and Historic Imagery
4
1a
Snow patterns…
Through a cooperative effort with University of
Alaska and Poker Flat Research Range, we plan
to utilize an unmanned aerial vehicle (UAV) to
collect remotely sensed data in precisely the same
locations to contrast efficiency, capability and quality
against our more traditional forms of survey.
When comparing the data collected via photo
interpretation (PI) to the data collected via aerial survey:
•PI yielded only 25% of the acreage
These results are a
compilation from both Mt.
•PI produced 25 times more discrete polygons
Edgecumbe and Peril Strait
1929 Navy photo
photos from Mt. Edgecumbe
2007 Satellite image
The overarching goals of the yellow-cedar
yellow-cedar decline program
program
are to better understand the cause of the decline problem,
and to use this knowledge and related information to
to
develop a conservation strategy and improved risk
maps for this valuable species.
species.
For more visit: www.yellow-cedar.net
Boeing’s Scan Eagle has a
10-foot wingspan, can fly for
20 hours on about 1-gallon
of gas and can be set up
with multiple imaging
systems.
With the drone’s low flying and multiple imaging
capabilities, we hope to improve upon our live cedar map.
Vertical and oblique aerial photos were
taken from Navy bi-wing seaplanes.
The reduction of persistent snow pack is the primary driver in
this forest decline. Snow can be also assessed through
retrospective remote image analysis of both vertical and
oblique imagery. We plan to compare temporal patterns of
snow, climate and decline.
Poster produced 12/31/2007
dwittwer@fs.fed.us (907) 586 7971