Brendan M.
a
Rogers ,
Ronald P.
b
Neilson ,
Ray
b
Drapek ,
James M.
b
Lenihan ,
and John R.
b
Wells
a
Department of Forest Science, Oregon State University, Corvallis, OR 97331
B USDA Forest Service Pacific Northwest Research Station, Corvallis, OR 97331
CSIRO MK3
(cool and wet)
Miroc 3 Medres
(hot and wet)
(cool and wet)
800
600
500
y = 169.83x
R² = 0.9172
p < 0.05
400
300
200
Change in TEC
100
2
0
1
0
1
Hadley CM3
(hot and dry)
2
3
4
5
mean annual temperature increase (ºC)
Precipitation
differences
(mm(H2O)/month)
+4.1
+4.0
+17.6
-6.6
+9.0
0
-1
CSIRO B1 CSIRO A1B CSIRO A2
Miroc B1
Miroc A2
Hadley B1 Hadley A1B Hadley A2
-2
-3
-5
2500
2000
Historical
CSIRO A1B
Hadley A2
1500
The graphs below illustrate the dynamics of future production drivers. In the Hadley A2 scenario,
net primary productivity (npp) reaches higher spring and fall peaks compared to the historical
period because higher temperatures (tmp) overlap with winter precipitation (ppt). In the summer,
however, decreased precipitation and higher vapor pressure deficit (vpd) deplete soil water, and
npp drops to near zero.
Historical
Hadley A2
1000
41
42
43
44
45
46
degrees latitude
47
48
49
50
*Results compare mean future (2070-2099) to mean historical (1971-2000) conditions
Cascade mountains lose subalpine forests,
maritime evergreen needle-leaf forests, and
shrublands to temperate evergreen needle-leaf forests and grasslands. The subalpine forests are lost because of higher
temperatures. Maritime forests transition into temperate forests because of larger temperature ranges. Shrublands are
converted to grasslands due to higher fire frequencies and intensities. As evidence of this, in the Okanagan ecoregion, which
contains the domain’s shrublands, mean burned area increased from 211 km2•yr-1 to 479 km2•yr-1 and mean biomass burned
increased from 23 g(C) •m-2•yr-1 to 62 g(C) •m-2•yr-1, averaged across all scenarios.
Future maps display results for the A2 emission scenario.
ppt
vpd
tmp
soil water
npp
Historical (1971 - 2000)
Jan Feb
Mar
Apr May
Jun
Jul
Aug
Sep Oct
Jan Feb
Nov Dec
Mar
Apr May
Jun
Jul
Aug
Sep Oct
Nov Dec
*Units are left out for illustrative purposes. Both graphs are on the same scale, and display average monthly
values for the historical (1971-2000) and future (2070-2099) periods in the North Cascades ecoregion.
Ecosystem Extent by Model
Precipitation and Summer VPD Predict Carbon
Historical
CSIRO
Miroc
Hadley
80
60
40
Future (2070 - 2099)
CSIRO MK3
(cool and wet)
Miroc 3 Medres
(hot and wet)
Hadley CM3
(hot and dry)
Because summer drought stress and
spring and fall temperatures have a large
influence on production, changes in total
ecosystem carbon (TEC) can be predicted
within each ecoregion by a linear function
of changes in yearly temperature and
summer vapor pressure deficit. Shown
here is a graph of the statistical fit for the
North Cascades.
Regression Model:
∆TEC = -0.92 + 0.84*
∆(yearly tmp) – 0.0058 ∆(summer vpd)
20
p < 0.05
*units are kg(C)/m2 for TEC, °C for tmp, and Pascals for
vpd
1.5
1
MC1 change in TEC (kg(C)/m2)
100
thousands of km2
Miroc A1B
-4
3000
mean ecotone (meters)
+2.9
(hot and dry)
700
Subalpine Ecotone Elevation by Latitude
Temperature
differences (°C)
(hot and wet)
TEC differences
(kg(C)/m2)
kg(C)/m2
We downscaled climate data output from three GCMs (Miroc 3 Medres, Hadley CM3, and
CSIRO MK3), each run at three SRES emission scenarios (B1, A1B, and A2), to develop
nine future climates on an 800-meter resolution grid. Shown below are differences
between future (2070-2099) and historical (1971-2000) mean monthly temperatures and
precipitation for the A2 scenario. Each scenario’s domain mean overlays its map.
Changes in total ecosystem carbon (TEC) are the product of competing climate-induced drivers.
An extended growing season increases the synchrony between temperature-dependent
productivity and winter precipitation. However, increased vapor pressure deficit and a longer
growing season increase drought stress in late summer. TEC changes for the A2 scenario are
shown below.
CSIRO MK3
Miroc 3 Medres
Hadley CM3
Subalpine Ecotone Elevation Increase
900
mean elevation increase (meters)
This study assesses the range of potential
vegetation and carbon budget changes in the
Pacific Northwest Cascade mountain range due to
projected climate change in the 21st century. MC1,
a dynamic general vegetation model, was run under
nine future climate scenarios. We analyze the
mechanisms causing changes in vegetation
distributions, subalpine ecotone elevations, and
carbon loads.
Subalpine ecotone elevation is a function of the magnitude and seasonality of
temperature increases, which determine growing degree days, and local topography and
lapse rates. Given this, increases in ecotone elevations display a strong linear
relationship to mean annual temperature increases.
0.5
0
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
-0.5
-1
-1.5
-2
-2.5
Regression-predicted change in TEC (kg(C)/m2)
0
Subalpine
Forest
Maritime
Evergreen
Needleleaf
Forest
Temperate
Evergreen
Needleleaf
Forest
Temperate
Shrubland
*Results are averaged across all three emission scenarios
Temperate
Grassland
Acknowledgements
This work was funded by the U.S.D.A. Forest Service, and made possible through the generous support of the MAPSS
team. For their insights and help with code fixes, we gratefully thank Dominique Bachelet and David Conklin .
Historical climate data was provided by the PRISM group, led by Christopher Daly. Help in downscaling climate data
was provided by Maureen Mcglinchy and Lauren Hahl.