CTCD Science Overview 20045

1
Making C flux calculations interact with
satellite observations of land surface properties
Shaun Quegan and friends
2
Global Carbon Data Assimilation System
Ciais et al. 2003 IGOS-P Integrated Global Carbon
Observing Strategy
3
Terrestrial Component
Water components SWE soil moisture
4
The SDGVM carbon cycle
Fire
Mortality
GROWTH
Thinning
NBP
Disturbance
Biomass
LEACHED
5
The Structure of a Dynamic Vegetation Model
Parameters
Climate

Sn
Sn1
DVM
Soil texture
Testing
Processes
6
EO interactions with the DVM
Parameters
7
Matching of concepts
Real world
S
Primary observation
Model
Model
Derived parameter
8
MODIS LAI/fAPAR biome Landcover 2000
MODIS/IGBP Landcover 2000
 
MODIS/UMD Landcover 2000
9
GLC2000 (SPOT-VGT)
CEH LCM2000
10
Scale effects on flux estimates (GLC-LCM)
GPP
NPP
NEP
Difference in annual predicted fluxes for GB,
1999. GLC LCM.
11
Lessons 1

• Land cover matters.
• Subjective land cover may be more useful than
  objective land cover.
• Scale matters.
• Can we do this better?

 
12
The SDGVM budburst algorithm
T0
Start of budburst
13
Data

• SPOT-VEG budburst 1998, 2000-02 0.1o
• Ground data Komarov RAS, dates of bud-burst at 9
  sites in the region.
• Temperature data ERA-40, 1.125o
• GTOPO-30 DEM
• Land cover GLC2000

14
The Date of budburst derived from minimum NDWI
(VGT sensor, 2000) N. Delbart, CESBIO
Day of year
15
Variability in optimising coefficients
16
Application of model to entire boreal regions
Model 1985
Model 2002
EO 2002
EO 1985
17
Comparison of ground data with calibrated model
18
Impact on Carbon Calculations
1 day advance NPP increases by 10.1 gCm-2yr-1 15
days advance 38 bias in annual NPP
Observations
Carbon Calculation
Dynamic Vegetation Model
Phenology model
Picard et al.,GCB, 2005
19
Comparison Model-EO RMSE
Model needs to be region specific, here include
chilling requirement ?
20
Lessons 2

• A simple 2-parameter spring warming model gives a
  good fit between model and EO data
• RMS differences between model, VGT data and
  ground data are 6.5 days.
• Ground data are crucial in investigating bias.
• Model failures are identifiable.
• Noise errors in NPP estimates are 8. Bias
  effects are 2.2 per day.
• Biophysical content of the parameters is low.

 
21
SDGVM module driven by climate data
Snow water equivalent (SWE)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
CTCD Comparison model and EO ( IIASA snow map)
Snow Water Equivalent (mm) 01/97 SSM/I
SDGVM using ECMWF
IIASA maximum snow storage
26
Lessons 3

• The physical quantity inferred from the EO data
  is almost certainly not what it is called.
• The problem here is making the model and the EO
  data communicate. Until communication is
  established, the data cannot be used to test or
  calibrate the model.

 
27
Severity of disagreement AVHRR/SDGVM
1998
r gt 0.497 OR r.m.s.e lt 0.2
r lt 0.497 AND r.m.s.e gt 0.2
r lt 0.497 AND r.m.s.e gt 0.3
28
Severity of disagreement example
Mid Europe
29
Severity of disagreement example
SW China
30
Lessons 5

• The DVM as currently formulated only supports a
  simple observation operator. This allows
  meaningful estimates of time series of
  observables absolute values of the observables
  are of dubious value.
• These time series permit the model to be
  interrogated with satellite data, and model
  failures to be identified.

 
31
Detecting incorrect land cover
Crop class incorrectly set
Crop class correctly set
0.9
0.0
Pearsons product moment
Temporal correlation
32
Final remarks

• The link between satellite measurements and most
  surface parameters used by the C models (and how
  they are represented) is indirect.
• In many cases, the only viable source of
  information on surface properties is from
  satellites.
• The art is to find the right means of
  communication between the data and the models.

33
Environmental effects on coherence
Coherence of Kielder Forest, July 1995

• Measurements by radar satellites are sensitive to
  biomass, but
• only for younger ages
• weather dependent through soil and canopy
  moisture

34
Age Estimation Accuracy
Raw Coherence

• Small Spatial Scale
• Inter-stand variance
• Inter stand bias

Time
Kielder Forest
Kielder Forest
North South

• Large Scale
• Meteorology dominant

35
Estimating NEE with SAR

N(age)
coherence
NEE tc ha-1 y-1 -8 -4 0
4 8
0 10 20 30 40 50
60 70 Age (y)
0 5 10 15 20 25 30 35
40 Age (y)
age
Sensitivity range
36
Using SPA to model coherence
Observations Model with biomass
saturation information
Model Backscatter
SPA was used to predict canopy and soil moisture,
and coupled with a radar scattering model to
predict coherence. Also needed was the saturation
level of biomass, which had to be measured from
the data
37
Lessons 3
Here the carbon model is essential to interpret
the data and its variation.
 
38
UK Forest NEE Calculations 1995
39
MODIS Burned Area
Russian Federation 500m burned areas 1 month 2002
40
MODIS Active Fires ( FRP)
Russian Federation 1km active fires 1 month 2002