Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
LAND PRODUCT VALIDATIONSUBGROUP REPORT
F. Baret, S. Garrigues & J. Morissette
WGCV-28 Sanya February 2008
WGCV-28 Sanya February 2008 2/25
OutlineOutline
• accomplishments since last WGCV
• relevance to the CEOS/IP or GEO task
• current/future challenge
• requested agency commitments.
WGCV-28 Sanya February 2008 3/25web curator: Jaime Nickeson, NASA GSFC
LPV web site maintenanceLPV web site maintenance
Thanks to Jaime Nickeson and Jeff Morissette
WGCV-28 Sanya February 2008 4/25
Previous workshopsPrevious workshops1) 7-8 June 2001 LAI Intercomparison
ESA Frascati, Italy2) 23-24 October 2002 Land Product Validation Workshop on Surface Albedo
Boston MA USA3) 16 August 2004 EOS LAI Intercomparison Activity Results
Missoula, MT USA4) 27-28 October 2005 Global Vegetation Continuous Fields Validation Workshop
Brookings, SD USA5) 27-28 April 2005 LPV workshop on albedo
Vienna, Austria6) 7 August 2006 LPV workshop on long-term VI record
Missoula Montana7) 8-10 August 2006 Long term global monitoring using moderate resolution satellites
Missoula Montana8) 23 March 2007 LPV workshop on LAI and fAPAR products
Davos, Switzerland
WGCV-28 Sanya February 2008 5/25
LAI validation exercise: refinements and outreachLAI validation exercise:
refinements and outreach
• Discussed during Missoula and Davos meetings• 2 papers published
– Weiss, M., F. Baret, S. Garrigues, R. Lacaze, and P. Bicheron. 2007. LAI, fAPAR and fCover CYCLOPES global products derivedfrom VEGETATION. part 2: Validation and comparison withMODIS Collection 4 products. Remote sensing of Environment, 110:317-331.
– Garrigues, S., R. Lacaze, F. Baret, J. Morisette, M. Weiss, J. Nickeson, R. Fernandes, S. Plummer, N. V. Shabanov, R. Myneni, and W. Yang. 2008. Validation and Intercomparison of Global LeafArea Index Products Derived From Remote Sensing Data. Journal of Geophysical Research, accepted.
WGCV-28 Sanya February 2008 6/25
Products consideredProducts considered
Clumping representation Product name GSD Temporal
Sampling
Algorithm Parameterization
Plant/Shoot Canopy Landscape
Seasonal smoothing
References
CYCLOPES-V3.1 1/112° 10 days RTM 1D (neural network) Global No No Yes No Baret et al., 2007 ECOCLIMAP 1/120° 1 month Empirical (based on NDVI
variations) Vegetation type Yes Yes No No Masson et al., 2003
GLOBCARBON 1/11.2° 1 month Model derived VI-LAI relationship
Vegetation type Yes Yes No Yes Deng et al., 2006
MODIS-C4 1km 8 days Main Alg: RTM 3D ( LUT) Backup: LAI-NDVI empirical relationship
Vegetation type Yes Yes Yes No Knyazikhin et al., 1998;Myneni et al., 2002
CCRS 1km 10 days Empirical VI-LAI relationship
Vegetation type Yes Yes Up to ~1ha Yes Fernandes et al., 2003
Consider 10x10 km² resolution1 month time step
WGCV-28 Sanya February 2008 7/25
0 5 10 15 20 25 30 35
-60
-40
-20
0
20
40
60
component fraction (%)
latit
ude
(°)
GrassCropsEvergreen broadleafEvergreen NeedleleafDeciduous BroadleafBare surface
Intercomparison over CEOS-BELMANIP
D
D
D DDD D
D
D
D
D D
D
D
DD
D
D
DDD
D
D
D
D D
D
DD
D
D
DD
D
D
DD
D
D
DD
D
D DDDDD
D
D
DD
D
D
DDD
D
DDD D
D
DD DD D DD
DDD
DD D
DDD DD D
D
D
DD
DD
D
DD
DD
D
FFF
F
FF
F
FF
F
FFFF
F F
F
FFFF
FFF FFF F F
FFFFF FFF
F
F
FF
F
FF
FF
FF
F
F
F
FFFFFF
FF
F FF
F
FFFFFFFFF
FFFFF
F
F
FFFF
FFFFFF FFFF
FF FFFFFF
F
FFF
F
FF
FFF F
F
FFF
F
F
FF
FFF
FFFFFF
FF
FFFFFF FF F
FFFFF FFFFFF
FF
F
FF FF F
F F
FF
F FFFFF
AA
A
A
A
A
A
A
AA
AA
AAA
AA
A
A
A
A
A
A
AA
A
A
A
A
A
A
AA
A A
A
AA
CC C
CC C
C
C
C C CC
C
C CCC
CCCC
CC
CC C C
C
C
CC C
CC
CCCC
C
CC
C
C CC CC
C
C C
CCC C
C CC C
CC
C
C C
C
C
C
C
CC
C
C
C
C
C
C
C
-180 -150 -120 -90 -60 -30 0 30 60 90 120 150
90
60
30
0
-30
-60
GrassCropsEvergreen broadleafConifersDeciduous BroadleafBare surface
397 sites representing the variability over surface types and latitudes
Baret, F. et al., 2006. Evaluation of the representativeness of networks of sites for the global validation and inter-comparison of land biophysical products. Proposition of the CEOS-BELMANIP. IEEE Transactions on Geoscienceand Remote Sensing, 44(7: special issue on global land product validation): 1794-1803.
WGCV-28 Sanya February 2008 8/25
IntercomparisonIntercomparison
0 2 4 6 8 100
20
40
60
80
100
cum
ulat
ive
freq
uenc
y (%
)
LAI values
January
CYCLOPESECOCLIMAPGLOBCARBONMODIS
0 2 4 6 8 100
20
40
60
80
100
cum
ulat
ive
freq
uenc
y (%
)LAI values
July
CYCLOPESECOCLIMAPGLOBCARBONMODIS
Statistical distribution: large differences between products
WGCV-28 Sanya February 2008 9/25
IntercomparisonIntercomparison
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.51, RMSE=1.04
CYCLOPES
EC
OC
LIM
AP
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.65, RMSE=0.56
CYCLOPESG
LOB
CA
RB
ON
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.71, RMSE=0.57
CYCLOPES
MO
DIS
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.40, RMSE=1.25
ECOCLIMAP
GLO
BC
AR
BO
N
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.42, RMSE=0.91
ECOCLIMAP
MO
DIS
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.52, RMSE=0.82
GLOBCARBONM
OD
IS
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.41, RMSE=1.42
CYCLOPES
EC
OC
LIM
AP
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.62, RMSE=0.98
CYCLOPES
GLO
BC
AR
BO
N
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.62, RMSE=1.15
CYCLOPES
MO
DIS
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.52, RMSE=1.59
ECOCLIMAP
GLO
BC
AR
BO
N
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.43, RMSE=0.95
ECOCLIMAP
MO
DIS
0 1 2 3 4 5 6 7 8 9 100
1
2
3
4
5
6
7
8
9
10r2=0.52, RMSE=1.32
GLOBCARBON
MO
DIS
Scatterplots: large differences depending on productsand vegetation type
WGCV-28 Sanya February 2008 10/25
IntercomparisonIntercomparison
-35 -30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30 35Latitude
0
1
2
3
4
5
6
7
8
LA
I Ju
ly-2
001
CYCLOPESECOCLIMAPGLOBCARBONMODIS
Op
en S
hru
bla
nd
Bare LandSah
elia
nS
avan
na
Sah
elia
nW
oo
od
lan
d
Eq
uat
ori
alA
fric
anF
ore
st
Eq
uat
ori
alW
oo
ded
Gra
ssla
nd
Su
btr
op
ical
Wo
od
edG
rass
lan
d
Bar
e L
and
Kal
ahar
iB
ush
lan
d
Mer
idio
nal
Afr
ican
Gra
ssla
nd
Bu
shla
nd
Eve
rgre
en W
oo
dla
nd
Spatial distribution:Similar patterns butDifferences in amplitude and smoothness
Africa (25˚ long. East)
WGCV-28 Sanya February 2008 11/25
IntercomparisonIntercomparison
Spatial distribution:Differences in valuesBut Similar patternsexcept in specific areas
maps of differences CCRS-Product over Canada
WGCV-28 Sanya February 2008 12/25
IntercomparisonIntercomparison
J DJ D J D0
1
2
3
4SEVI, USA, Lat=34.35, Lon=−106.69
2001 2002 2003month
LAI
J DJ D J D0
1
2
3
4Larzac, France, Lat=43.94, Lon=3.12
2001 2002 2003month
LAI
J DJ D J D0
1
2
3
4Gourma, Mali, Lat=15.32, Lon=−1.56
2001 2002 2003month
LAI
J DJ D J D0
1
2
3
4
5
6
7Nezer, France, Lat=44.57, Lon=−1.05
2001 2002 2003month
LAI
J DJ D J D0
1
2
3
4
5
6
7Thompson, Manitoba, Canada, Lat=56.05, Lon=−98.16
2001 2002 2003month
LAI
J DJ D J D0
1
2
3
4
5
6
7Flakaliden, sweden, Lat=64.11, Lon=19.47
2001 2002 2003month
LAI
J DJ D J D0
1
2
3
4
5
6
7HARV, USA, Lat=42.53, Lon=−72.17
2001 2002 2003month
LAI
J DJ D J D0
1
2
3
4
5
6
7Walker Branch, TN, USA, Lat=35.96, Lon=−84.29
2001 2002 2003month
LAI
J DJ D J D0
1
2
3
4
5
6
7Larose2, Canada, Lat=45.38, Lon=−75.17
2001 2002 2003month
LAI
Temporal distribution: similar patterns (except ecoclimap)But differences in continuity (smoothness, gaps)
Grassland
Needle leafforest
Deciduousforest
WGCV-28 Sanya February 2008 13/25
Direct validation: Comparison with ground measurements:bottom-up approach
Direct validation: Comparison with ground measurements:bottom-up approach
High spatial resolutionimage (SPOT/TM/ASTER …)
Individual measurements
Value at the ESU level
Value(s) at the site level
Global validation
Averaging
Transferfunction
Scatter plot
50-100 sites
10-100 measurements/ESU
20-100 ESUs/site
D
D
D DDD D
D
D
D
D D
D
D
DD
D
D
DDD
D
D
D
D D
D
DD
D
D
DD
D
D
DD
D
D
DD
D
D DDDDD
D
D
DD
D
D
DDD
D
D
DD D
D
DD DD D DD
DDD
DD D
DDD DD D
D
D
DD
DD
D
DD
DD
D
FFF
F
FF
F
FF
F
FFFF
F F
F
FFFF
FFF FFF F F
FFFFF FFF
F
F
FF
F
FF
FF
FF
F
F
F
FFFFFF
FF
F FF
F
FFFFFFFFF
FFFFF
F
F
FFFFFFFFFF FFFF
FF FFFFFF
F
FFF
F
FF
FFF F
F
FFF
F
F
FF
FFF
FFFFFF
FF
FFFF FF FF F
FFFFF FFFFFF
FF
F
FF FF FF F
FF
F FFFFF
AA
A
A
A
A
A
A
AA
AA
AAA
AA
A
A
A
A
A
A
A
AA
A
A
A
A
A
AA
A A
A
AA
CC C
CC C
C
C
C C CC
C
C CCC
CCCC
CC
CC C C
C
C
CC C
CC
CCCC
C
CC
C
C CC CC
C
C C
CCC C
C CC C
CC
C
C C
C
C
C
C
C C
C
C
C
C
C
C
C
Medium resolutionproducts to be validated
WGCV-28 Sanya February 2008 14/25
Direct Validation: resultsDirect Validation: results
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7
LAI reference map
EC
OC
LIM
AP
RMSE=1.36RMSE
u=1.38
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7
LAI reference mapG
LOB
CA
RB
ON
RMSE=1.15RMSE
u=1.15
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7
LAI reference map
CY
CLO
PE
S
RMSE=0.70RMSE
u=0.70
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7
LAI reference map
MO
DIS
RMSE=1.28RMSE
u=1.27
Decid. Broad. For.Ever. Needle. ForestEver. Broad. ForestCropsGrassMixed Forest (DBF+ENF)
WGCV-28 Sanya February 2008 15/25
SynthesisSynthesis
Criteria CYCLOPES ECOCLIMAP GLOBCARBON MODIS Global distribution + - + + Spatial consistency + - - + Temporal consistency
Missing data - + + - Interannual and
seasonal cycles + - - +
Magnitude of LAI High values - + + + Low values + + + -
Comparison to in-situ data
+ - - -
WGCV-28 Sanya February 2008 16/25
LAI Validation: conclusionLAI Validation: conclusion• Methods have been developed and published
• Morisette, J., F. Baret, and 28 co-authors. 2006. Validation of global moderate resolution LAI Products: a framework proposed within the CEOS Land Product Validation subgroup. IEEE Transactions on Geoscience and Remote Sensing, 44:1804-1817.
• Baret, F., J. Morissette, and 7 co-authors. 2006. Evaluation of the representativeness of networks of sites for the global validation and inter-comparison of land biophysical products. Proposition of the CEOS-BELMANIP. IEEE Transactions on Geoscience and Remote Sensing, 44:1794-1803.
• Methods have been applied and results published• Weiss, M., F. Baret, S. Garrigues, R. Lacaze, and P. Bicheron. 2007. LAI, fAPAR and
fCover CYCLOPES global products derived from VEGETATION. Part 2: Validation and comparison with MODIS Collection 4 products. Remote sensing of Environment, 110:317-331.
• Garrigues, S., R. Lacaze, F. Baret, J. Morisette, M. Weiss, J. Nickeson, R. Fernandes, S. Plummer, N. V. Shabanov, R. Myneni, and W. Yang. 2008. Validation and Intercomparison of Global Leaf Area Index Products Derived From Remote Sensing Data. Journal of Geophysical Research, accepted
• But…– Need updates with new products versions : on line validation exercise– Small number of validation sites: continuous increase of number of sites (new
sites and inclusion of processed archive sites) to get closer to stage 3 validation– Very little ‘continuous’ ground LAI measurements: PAR@METER devices– Revise BELMANIP: lack of interpixel homogeneity for some sites– Application to other products: fAPAR
WGCV-28 Sanya February 2008 17/25
Towards the development of virtualconstellation products
Towards the development of virtualconstellation products
• Objectives:– Development of consistent products from several
sensors to allow simple fusion• Framework:
– GEO/CEOS virtual constellation– GEOLAND_2
• Approach:– Evaluate the performances of neural networks when
trained with several inputs and outputs– Application to CYCLOPES/VEGETATION and MODIS
WGCV-28 Sanya February 2008 18/25
Products usedProducts usedNNT
Reflectance (MOD43B4)
LAI (MOD15A2)
Reflectance (CYCL L3a)
LAI (CYCL V3.1) resampling
spatial sampling 3 kmprojection sinusgeometric accuracy 50-140PSF ? ? ?-temporal sampling 16 8 10 10 16Temporal resolution 16 8 30 30 24-40
Red 620-670 610-680NIR 841-876 780-890SWIR 1628-1652 1580-1750Clumping plant/canopy landscape
1920m50 m 140m
MODIS CYCLOPES
1kmsinus
1/112°lat-lon
Original CYCLOPES (V3.1) and MODIS (collection 4) products
Resampling necessary for better consistency of productsFiltering over the 3x3 pixels: >50% good data
Biophysical algorithm diferences for LAI- MODIS: LUT computed from 3D RT model and tuned for 6 biome classes- CYCLOPES: NNT trained from 1D RT model across all classes
WGCV-28 Sanya February 2008 19/25
Temporal resolutionTemporal resolution
Differences in temporalresolution betweenMODIS and CYCLOPES
WGCV-28 Sanya February 2008 20/25
Sites samplingSites sampling
• 397 BELMANIPsites
• 2001-2003 period
• Filtering:– 38112 potential observations– only 7572 matches (20%)– 10% ‘outliers’ eliminated
D
D
D DDD D
D
D
D
D D
D
D
DD
D
D
DDD
D
D
D
D D
D
DD
D
D
DD
D
D
DD
D
D
DD
D
D DDDDD
D
D
DD
D
D
DDD
D
D
DD D
D
DD DD D DD
DDD
DD D
DDD DD D
D
D
DD
DD
D
DD
DD
D
FFF
F
FF
F
FF
F
FFFF
F F
F
FFFF
FFF FFF F F
FFFFF FFF
F
F
FF
F
FF
FF
FF
F
F
F
FFFFFF
FF
F FF
F
FFFFFFFFF
FFFFF
F
F
FFFFFFFFFF FFFF
FF FFFFFF
F
FFF
F
FF
FFF F
F
FFF
F
F
FF
FFF
FFFFFF
FF
FFFF FF FF F
FFFFF FFFFFF
FF
FFF FF F
F F
FF
F FFFFF
AA
A
A
A
A
A
A
AA
AA
AAA
AA
A
A
A
A
A
A
A
AA
A
A
A
A
A
AA
A A
A
AA
CC C
CC C
C
C
C C CC
C
C CCC
CCCC
CC
CC C C
C
C
CC C
CC
CCCC
C
CC
C
C CC CCC
C C
CCC C
C CC C
CC
C
C C
C
C
C
C
C C
C
C
C
C
C
C
C
-180 -150 -120 -90 -60 -30 0 30 60 90 120 150
90
60
30
0
-30
-60
WGCV-28 Sanya February 2008 21/25
Consistency between original productsConsistency between original products
OK for reflectances (biaises between 3 to 7 %)
Large scattering for LAI due to products definitionAssumption on architectureand algorithm principles
WGCV-28 Sanya February 2008 22/25
Neural network training processNeural network training process
MODrhoMODlai0.502 (0.558)
CYCrhoMODlai0.525 (0.609)
MODIS LAI
MODrhoCYClai0.297 (0.305)
CYCrhoCYClai0.119a (0.129b)
CYCLOPES LAIMODIS reflectancesCYCLOPES reflectances
θs
B2
B3
SWIR
Norm
Norm
Norm
Norm
NormS
S
S
S
S
L Var
Red
NIRLAI
Four combinations investigated
aRMSE per vegetation class bRMSE all vegetation classes
Good capacity of neural networks to ‘learn’ any algorithm
WGCV-28 Sanya February 2008 23/25
Satistical distribution across biomesSatistical distribution across biomes
Good consistency when trained with same LAI but different reflectance sourcePoor consistency between CYCLOPES and MODIS original LAI products values
WGCV-28 Sanya February 2008 24/25
Results: Temporal profilesResults: Temporal profiles
Products behavior depends mainly on the LAI from which it derives
WGCV-28 Sanya February 2008 25/25
Direct ValidationDirect Validation
WGCV-28 Sanya February 2008 26/25
Towards virtual constellation products: CONCLUSION
Towards virtual constellation products: CONCLUSION
• Innovative approach allowing easy fusion of observations coming from several sensors
• Importance of the training data base (LAI)• Applies to any product (even albedo!)• No need for absolute reflectance calibration or strong
spectral consistency between sensors: just temporal and spatial stability!
• Application over instantaneous observations (including or not atmosphere)
• Need good geometrical consistency• First implementation in GEOLAND_2 project (EC) for
Long Time Series (as opposed to Near Real Time)• Results published:
– Verger, A., F. Baret, and M. Weiss. 2008. Efficiency of neural networks for consistent calibration of LAI products from input reflectance coming from several sensors: Application to CYCLOPES/VEGETATION and MODIS data. Remote Sensing of Environment, accepted for publication.
WGCV-28 Sanya February 2008 27/25
Contribution to GEO TasksContribution to GEO TasksSeveral tasks share same contributions from LPV.
Area Task # Task Title Proposition for potential contribution from LPV LPV achievements Contact ReferencesAgriculture AG-06-04 Forest
Mapping and Change Monitoring
1) Recommendations for Evaluation and Accuracy Assessment of Global Land Cover Maps;
2) Proposition of using biophysical products for change detection.
1) LPV with GOFC-GOLD contributed to this task by encouraging coordinated developments and providing the framework for evaluation of Global Land Cover maps. Documents were written that report methodology and results for the validation of land cover maps.
2) To be discussed and tested
Alan H. Strahler, Philippe Mayaux, LPV chair
1, 2, 3, 4
Agriculture AG-07-01 Improving Measurements of Biomass
1) provide accuracy evaluation of inputs to vegetation productivity models
2) provide recommendations for accuracy assessment of productivity, gross primary production and biomass
1) Biophysical variables such as LAI and fAPAR are key inputs to vegetation productivity models. LPV has been focusing on those variables, by (i) proposing a clear and consensus definition of these variables, (ii) describing departure of available products from the main definition, (iii) developed of a methodological framework for accuracy assessment of products, and (iiii). Results on evaluation of medium spatial resolution LAI products accuracy from ground measurements and intercomparison between available products.
2) Discussions started at Global Vegetation Monitoring meeting (Missoula 2006) and papers published in IEEE TGARS validation special issue.
LPV Chair, Sebastien Garrigues
5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
16, 17Climate CL-06-02 Key Climate
Data from Satellite Systems
Climate CL-06-03 Key terrestrial observations for Cli tData
ManagementDA-06-02 GEOSS
Quality Assurance Strategy
1) Provide a strategy and methods for quality assessment of land cover (with GOFC-GOLD) fire and biophysical products derived from satellite observations.
2) Maintain easy access to key information for the validation LPV.
3) Organizes meetings to define and discuss the methods and share results with the community.
1) A strategy has been defined for the validation of higher level products: land cover (with GOFC-GOLD) and fire and LAI and fAPAR biophysical products. Results show that stage 2 of the validation is achieved, but stage 3 (quantitative accuracy assessment representative of global conditions) is not yet reached.2) A web site is set up providing information on validation activities. Articles in peer reviewed journal have been published with results based on methods proposed within LPV.3) Workshops have been organized to define and discuss the methods and share results with the community.
LPV chair 2, 6, 7
21, 5, 8,9, 10, 11, 12, 13, 14, 15
21 22 23 18 5Data Management
DA-06-04 Data, Metadata and Products Harmonisation
1) Provide strategy for harmonization of global land cover mapping
2) Provide a strategy for harmonization through intercomparison of homologous biophysical products
1) A strategy for harmonization of global land cover has been defined, allowing intercomparison of classifications and maps.
2) A strategy for intercomparing biophysical products has been proposed. Preliminary results are available for LAI and fAPAR.
Martin HeroldLPV chair
2
5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15Data
ManagementDA-07-02 Global Land
Cover1) Recommendations for Evaluation and Accuracy Assessment of Global Land Cover Maps;
2) Proposition of using biophysical products for change detection.
1) LPV with GOFC-GOLD contributed to this task by encouraging coordinated developments and providing the framework for evaluation of Global Land Cover maps. Documents were written that report methodology and results for the validation of land cover maps.
2) To be discussed and tested
Alan H. Strahler, Philippe Mayaux, LPV chair
1, 2, 3, 4
Data Management
DA-07-03 Virtual Constellations
1) Propose a strategy for intercomparison of products derived from several sensors.2) Propose methods for merging products coming from several sensors3) Evaluate benefit of using virtual constellation products4) Define minimum requirements for inter-operability of sensors
1) A strategy for intercomparison of products was proposed with application to biophysical variables (LAI, fAPAR)2) Simple solutions proposed for merging products coming from similar sensors3) Early evaluation for albedo and BRDF products4) Not yet achieved
LPV Chair 5, 6, 7, 8
24, 25
24, 25
Disasters DI-06-09 Use of Satellites for Risk Management
1) Propose a strategy for intercomparison of products derived from several sensors.2) Propose methods for merging products coming from several sensors3) Evaluate benefit of using virtual constellation products4) Define minimum requirements for inter-operability of sensors
1) A strategy for intercomparison of products was proposed with application to biophysical variables (LAI, fAPAR)2) Simple solutions proposed for merging products coming from similar sensors3) Early evaluation for albedo and BRDF products4) Not yet achieved
LPV Chair 5, 6, 7, 8
24, 25
24, 25
Ecosystems EC-06-01 Integrated Global Carbon Observation (IGCO)
1) Provide uncertainties on products used to scale up local observations to region and scale
1) Uncertainties available for few products required in the scaling up process: land cover, albedo, LAI and fAPAR
LPV Chair 21, 5, 8,9, 10, 11, 12, 13, 14, 15
Ecosystems EC-06-02 Ecosystem Classification
1) Propose recommendations for Evaluation and Accuracy Assessment of Global Land Cover Maps;
2) Propose to use biophysical products for change detection.
1) LPV with GOFC-GOLD contributed to this task by encouraging coordinated developments and providing the framework for evaluation of Global Land Cover maps. Documents were written that report methodology and results for the validation of land cover maps.
2) To be discussed and tested
Alan H. Strahler, Philippe Mayaux, LPV chair
1, 2, 3, 4
Ecosystems EC-06-07 Regional Networks for Ecos stems
1) Development of validation of products for the monitoring of ecosystems at regional level
1) LPV developed a strategy for validation of land cover and biophysical products. Early results are available for land cover and LAI, fAPAR and albedo
LPV ChairA. StrahlerP. Mayaux
1, 2, 3,4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15
Ecosystems EC-07-01 Global Ecosystem Observation and Monitoring Network
1) Contribute to provide a global classification of ecosystems
2) Contribute to develop a global sampling scheme for ecosystem characterization and monitoring
3) Propose methods to up-scale local ground measurements to larger spatial domains
1) LPV with GOFC-GOLD contributed to this objective by encouraging coordinated development of consistent land cover mapping.2) LPV has proposed a global network of sites that samples vegetation types and conditions. It is based on existing thematic networks such as AERONET or FLUXNET and sites where ground measurements are collected for the direct validation of medium resolution biophysical products. Agencies agreed to provide extracts of medium resolution products they are in charge.3) In the framework of the validation of medium resolution products, up-scaling methods have been developped to extend over larger spatial domains a set of local ground measurements and qualify the spatial sampling used. These methods are based on high spatial resolution images
LPV chair 1, 2, 4
7
6, 26, 9, 10, 11, 12
Weather WE-06-02 Space-based Global Observing System for Weather
1) Provide accuracy statements for Essential Climate Variables (ECVs) as defined in the GCOS Implementation Plan for the terrestrial domain. Emphasis on land cover, Albedo, LAI, fAPAR, Fire and soil moisture products from medium resolution sensors.
2) Propose strategies to exploit in a consistent way historical satellite archive and derive long time series of products.
1) LPV with GOFC-GOLD contributed to validate global land cover maps. Similarly, LPV contributed in the validation of LAI, fAPAR and fire products.
2) LPV proposed strategies to build consistent long time series of biophysical variables
LPV chair 18, 8
20
LPV chair1) Provide accuracy statements for Essential Climate Variables (ECVs) as defined in the GCOS Implementation Plan for the terrestrial domain. Emphasis on land cover, Albedo, LAI, fAPAR, Fire and soil moisture products from medium resolution sensors.
2) Propose strategies to exploit in a consistent way historical satellite archive and derive long time series of products.
1) LPV with GOFC-GOLD contributed to validate global land cover maps. Similarly, LPV contributed in the validation of LAI, fAPAR and fire products.
2) LPV proposed strategies to build consistent long time series of biophysical variables
18, 8
20
WGCV-28 Sanya February 2008 28/25
Planed activitiesPlaned activities• Future meeting
– Albedo meeting AGU fall 2008– Soil moisture (SMOS) TBD– Temporal signature in remote sensing. January 2010??
• fAPAR products validation• Continue collecting ground validation• Test and exploit PAR@METER devices for
continuous fAPAR/LAI monitoring• Products PSF characterization• On line validation (CALVAL portal)• Populate calval portal (definitions, best practices)
WGCV-28 Sanya February 2008 29/25
RecommendationsRecommendations• Monitoring progress of previous recommandations
– Recommend agencies to support the continuity and expansion of product validation activities to be able to quantify the associated uncertainties and allow fusion between similar products
– Encourage agencies to prepare subsets of data/products for global product intercomparison activity as described by CEOS/WGCV/LPV
• Need more consistency in geometrical formats (grid/projection/datum)
• Need ressources for implementation of the ‘on linevalidation tool’ in the CAL/VAL portal