Carbon Flux

1
Carbon Flux Cumulative NEP of Disturbed Land
Mosaics in a Changing Climate
Jiquan Chen Kim Brosofske Asko Noormets Tom
Crow Soung Ryu Many others
Dept. of Earth, Ecological and Environmental
Science Landscape Ecology and Ecosystem Science
Laboratory
LEES
2
Current Questions

• What are the major obstacles to understanding
  landscape-scale flux processes when the landscape
  is composed of contrasting patches?
• How variable is NEC among patches of different
  types and ages?
• Our central hypothesis is that the cumulative
  NEP of a landscape is determined by the land
  mosaic -- the various ages and types of
  ecosystems present and their spatial arrangement.

Hypothesis
3
NEP Changes
Changes in NEP with age (a) and the age structure
of a hypothetical landscape (b) together
determine the cumulative NEP of the landscape (c)
4
Objective

• To enhance the understanding of landscape-level
  carbon exchange in disturbed land mosaics, taking
  into specific consideration age structure, patch
  type, and dynamics climate.

5
Conceptual framework for studying landscape-level
carbon flux and storage in disturbed land
mosaics, with explicit consideration of
disturbance regime and landscape structure,
including the area-of-edge influences (AEI).
6
Study Area
LandSAT ETM
7
Figure 1. Cover-type map of the study area
within the Chequamegon National Forest, WI
landscape. The labels on the map indicate the
location of the sampled patch types for soil
respiration measurements. CC clearcuts, JP
Jack pine, MH mature hardwoods, PB pine
barrens, RP red pine, YH young hardwoods
8

• Classification of the northern portion of the
  Chequamegon National Forest, Washburn Ranger
  District from 2001. Triangles indicate
  ecosystems containing Eddy-Covariance Towers

M. Red Pine
Y. Hardwood
Pine Barrens
Y. Red Pine
M. Hardwood
9
Methods

• A combination of direct measurements, remote
  sensing images, and ecosystem models.

10
Pine Towers
11
Results
Diurnal net exchange of CO2 at five forest
ecosystems in the Chequamegon National Forest, N.
Wisc. Data were collected using eddy-covariance
systems and WPL corrections were made.
12
Cumulative C flux of five ecosystems in CNF
between May and Sept., 2002.
Results
13
PnET Model
http//www.pnet.sr.unh.edu/
14
PnET A Generic Model
The Flow Chart of the Model
The GUI of PnET
15
PnET has been applied at many forests in USA
Oak-Maple
Douglas-fir
Lodgepole Pine
Northern Hardwood
Red Pine Oak Maple
Taiga Aspen White Spruce
Pine Mixed Hardwood
Slash Pine
16
parameterization

Site and soil variables Site and soil variables Site and soil variables Site and soil variables Photosynthesis Photosynthesis Carbon allocation Carbon allocation
Lat NetPsnMaxA Cfracleaf 0.45
WHC NetPsnMaxB CFracWood 0.5
Canopy variables Canopy variables Canopy variables Canopy variables BaseLeafRespFrac 0.1 CFracAllocA 130
K K K HalfSat CFracAllocB 1.92
LeafNCon LeafNCon LeafNCon PsnTmax 40 CFracRoot 0.45
FolReten FolReten FolReten PsnTop 20 LeafGRespFrac 0.25
LeafSpecWt LeafSpecWt LeafSpecWt PsnTmin 0 WoodGRespFrac 0.25
Water balance variables Water balance variables Water balance variables Water balance variables Water balance variables Water balance variables RootGRespFrac 0.25
VPDeffk VPDeffk PrecIntFrac RootMRespFrac 0.5
WUEConst WUEConst 10.9 10.9 FastFlowFrac 0.1 WoodMRespFrac 0.35
f f 0.04 0.04

17
Gross carbon exchange in a mature hardwood forest
in N. Wisconsin between May and September 2002.
GCE (g C m-2.dya-1)
18
Gross carbon exchange in a mature hardwood forest
in N. Wisconsin between May and September 2002.
GCE (g C m-2.dya-1)
Julian Day
19
Comparison of measured (NCERs) predicted GCE
(PnET model) for a mature hardwood forest in N.
Wisconsin in 2002.
Predicted GCE (g C m-2 day-1)
Minus Ra
11 line
Measured GCE (g C m-2 day-1)
20
Changes of predicted residuals with environmental
variables for a mature hardwood forest in N.
Wisconsin in 2002.
Predicted - Measured
21
Changes of predicted residuals with daily maximum
and minimum temperatures for a mature hardwood
forest in N. Wisconsin in 2002.
Residuals
Max Temp
Min Temp
22
Gross carbon exchange in a mature red pine forest
in N. Wisconsin between May and September 2002.
GCE (g C m-2.dya-1)
Julian Day
23
Gross carbon exchange in a mature red pine forest
in N. Wisconsin between May and September 2002.
GCE (g C m-2.dya-1)
Julian Day
24
Comparison of measured (NCERs) predicted GCE
(PnET model) for a mature red pine forest in N.
Wisconsin in 2002.
Predicted GCE (g C m-2 day-1)
11 line
Measured GCE (g C m-2 day-1)
25
Changes of predicted residuals with environmental
variables for a mature red pine forest in N.
Wisconsin in 2002.
Julian Day
PAR
Predicted - Measured
Max Temp
Min Temp
26
Changes of predicted residuals with daily maximum
and minimum temperatures for a mature red pine
forest in N. Wisconsin in 2002.
Residuals
Max Temp
Min Temp
27
Changes of predicted residuals with daily maximum
and minimum temperatures for a young hardwood
forest in N. Wisconsin in 2002.
Residuals
Max Temp
Min Temp
28
Changes of predicted residuals with daily maximum
and minimum temperatures for a young red pine
forest in N. Wisconsin in 2002.
Residuals
Max Temp
Min Temp
29
Changes of predicted residuals with daily maximum
and minimum temperatures for a pine Barrens in N.
Wisconsin in 2002.
Residuals
Max Temp
Min Temp
30
Comparison of measured (NCERs) predicted GCE
(PnET model) for a young hardwood forest in N.
Wisconsin in 2002.
Predicted GCE (g C m-2 day-1)
Measured GCE (g C m-2 day-1)
31
Comparison of measured (NCERs) predicted GCE
(PnET model) for a young red pine forest in N.
Wisconsin in 2002.
Predicted GCE (g C m-2 day-1)
Measured GCE (g C m-2 day-1)
32
Comparison of measured (NCERs) predicted GCE
(PnET model) for a pine barrens in N. Wisconsin
in 2002.
Predicted GCE (g C m-2 day-1)
Measured GCE (g C m-2 day-1)
33
(No Transcript)
34
Mature red pine
Mature hardwood
11 line
Predicted GCE (g C m-2 day-1)
Young hardwood
Young red pine
Pine barrens
Measured GCE (g C m-2 day-1)
35
Gross carbon exchange in a mature hardwood (left)
and red pine (right) forest in N. Wisconsin
between May and September 2002. The differences
are highly correlated with daily min and max
temperature.
GCE (g C m-2.dya-1)
Residuals
Residuals
36
Discussion
Exploring the unique biophysical controls of C
flux is essential for an mechanistic
understanding and predictions of the differences
and their changes over time.
37
Questions?
38
Mobile Unit Towers
Landuse Scenarios
GCM Predictions

(flux,LAI, etc.)
Land Cover
Climate
Climate
Land Cover
Image process comparison
Surface
RS
Surface
RS
Data
Data
Data
Data
parameterization
parameterization
spatialization
inversion
PnET SiB2
New Model
regional
Flux validation
feedback
CO
CH
Flux
CO
CH
Flux
2
4
2
4
Figure 1. Proposed framework for predicting
changes of CO2 and CH4 fluxes in five Midwest
states (WI, IL, IN, OH, and MI) under various
climate and land use scenarios.
39
Fig. 2. Land cover and locations of field
campaigns for predicting regional CO2 and CH4
fluxes. The region is dominated by forests,
croplands, pasture land (and wooded area), and
growing urban areas. Data collected at the 12
AmeriFlux towers is shared and publicly
available. A new flux tower in Toledo and 3
mobile systems will be deployed to sample 9-12
additional ecosystems. A Trace Gas Analyzer
TGA-100 will be added to the eddy-covariance
systems for measuring CH4 flux (with a circle).