Post on 27-Jun-2020
1
The CEOS Leaf Area Index Inter-comparison as aprototype activity
Global Vegetation Workshop, University of Montana
2
Global Vegetation Workshop, University of Montana
Outline
1)1) Background: CEOSBackground: CEOS
2)2) Global validation and inter-Global validation and inter-comparisoncomparison approachapproach
3)3) Inter-Inter-comparisoncomparison of: GLOBCARBON, CYCLOPES, MODIS, of: GLOBCARBON, CYCLOPES, MODIS,
ECOCLIMAP and CCRS global ECOCLIMAP and CCRS global datasetsdatasets
4)4) ConclusionsConclusions
3
Background (1/3):Leaf Area Index Product
• Leaf Area Index:
• Key surface characteristic for land surface modeling
• Essential Climate Variable as defined by GCOS
• LAI definition: half the total developed area of leaves per unit ground horizontal
surface area (Chen and Black, 1992)
Global Vegetation Workshop, University of Montana
4
Global Vegetation Workshop, University of Montana
Background (2/3)Rationale for validation and inter-comparison of LAI
5
Global Vegetation Workshop, University of Montana
• CEOS/WGCV subgroup created in 1999 to
Background (3/3)Land Product Validation group
6
Method (1/5)The BELMANIP Global Network of Sites
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
DDD
D
DD DD D DD
DDD
DD D
DDD DD D
D
D
D
D
DD
D
DD
DD
D
FFF
F
FF
F
FF
F
FFFFF F
F
FFFF
FFF FFF F F
FFFFFFF
F
F
F
F
FF
FF
FF
FF
F
F
F
FFFFFFFF
FFF
F
FFFFFFFFF
FFFFF
F
F
F
FFF
FFFFFF FFFF
FF FFFFFF
F
FFF
F
FF
FFF F
F
FFF
F
F
FF
FFF
FFFFFF
FF
FFFFFF FF F
FF
FFF FFFFFFFF
F
FF FF F
F F
F
FF FF
FFF
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 C
CC
CCCC
CC
CC C C
C
C
CC C
C
C
CCCC
C
C
CC
C CC C
CC
C C
CCC
CC C
C CCCC
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
Bare Water DBF ENF EBF Crops Grass
•representative sampling of global land surface types
•about 400 sites from several networks: direct validation sites (D: BIGFOOT, VALERI… )
AERONET (A), FLUXNET (F)…
Global Vegetation Workshop, University of Montana
7
0 5 10 15 20 25 30 35
-60
-40
-20
0
20
40
60
component fraction (%)
latitude (°)
GrassCropsEvergreen broadleafConifersDeciduous BroadleafBare surface
GrassCropsEvergreen broadleafEvergreen NeedleleafDeciduous BroadleafBare surface
0 5 10 15 20 25 30 35
-60
-40
-20
0
20
40
60
component fraction (%)
latitude (°)
GrassCropsEvergreen broadleafEvergreen NeedleleafDeciduous BroadleafBare surface
Global surface type distribution BELMANIP surface type distribution
76 sites were added to FLUXNET, AERONET and Direct Validation sites to
better sample latitude, longitude and surface types (grass, bare surface,
evergreen broadleaf forest)
ECOCLIMAP global classification (Masson et al, 2003)
Method (2/5)The BELMANIP Global Network of sites
Global Vegetation Workshop, University of Montana
8
Global Vegetation Workshop, University of Montana
Method (3/5): Direct Validation approach
Accuracy Assessment
LAI Field
data
High resolution
imagery
High resolution
LAI map Moderate resolution
LAI
Transfer
function
Aggregate&
Relate
9
Global Vegetation Workshop, University of Montana
Method (4/5)High Spatial Resolution LAI Map Database
10
Global Vegetation Workshop, University of Montana
Method (5/5): LAI inter-comparison approach
Projection&
resampling
Extractionover
BELMANIP sites
Inter-comparison metrics:Map of difference
Global histogram
Transect
Product/product scatter plot
11
Global Vegetation Workshop, University of Montana
Results: Products investigated
no
yes
yes
yes
yes
Total
LAI
yes
yes
yes
no
yes
Woody
correction
yesClimatology from global
climate distribution and
NOAA AVHRR time series
ECOCLIMAP
(Masson, et al.,2003)
yesSemi-empirical (SR, ISR,
BRDF correction, land use
specific retrieval, temporal
filter)
CCRS
(Fernandes, et al., 2003)
yesRadiative Transfer Inversion
by LUT
MODIS
(Myneni et al., 2002)
noRadiative Transfer Inversion
by neural network
CYCLOPES
yesSemi-empirical relationship
(SR,RSR VI, BRDF
Correction and temporal
filter)
GLOBCARBON
(Plummer et al)
Clumping
correction
AlgorithmProduct
12
Global Vegetation Workshop, University of Montana
Results: Global maps (Aug 2002)
13
0 2 4 6 80
0.2
0.4
0.6
0.8
1
Cumulative frequency (
LAI values
01/2002
Global Vegetation Workshop, University of Montana
Results: Global histogram
0 2 4 6 80
0.2
0.4
0.6
0.8
1
Cumulative frequency (
LAI values
07/2002
14
Global Vegetation Workshop, University of Montana
Results: Map of differences over CanadaJuly 2002
15
Global Vegetation Workshop, University of Montana
Results: Map of differences over Canada
16
Global Vegetation Workshop, University of Montana
Results: Map of differences over CanadaJuly 2002
17
Global Vegetation Workshop, University of Montana
Results:Transect over Africa (25˚ est)
18
Global Vegetation Workshop, University of Montana
Results:Time series over crop sites
Jan Dec Jan Dec0
0.5
1
1.5
2
2.5
3
3.5
4Crops (Fundulea, Romania)
2002 2003month
LA
19
Global Vegetation Workshop, University of Montana
Results: Time series over grassland
Jan Dec Jan Dec0
0.5
1
1.5
2
2.5
3
3.5
4Grass (Larzac, France)
2002 2003month
LAI
20
Global Vegetation Workshop, University of Montana
Results: Time series over DBF
Jan Dec Jan Dec0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5Decid. Broad. For. (Larose2, Canada)
2002 2003month
LAI
21
Global Vegetation Workshop, University of Montana
Results: Time series over ENF
Jan Dec Jan Dec0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5Ever. Needle. Forest (NOBS, Canada)
2002 2003month
LAI
22
Global Vegetation Workshop, University of Montana
Results: Time series over ENF
Jan Dec Jan Dec0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5Ever. Needle. Forest (WatsonLake, Yukon, Canada)
2002 2003month
LAI
23
Global Vegetation Workshop, University of Montana
Results: Time series over EBF
Jan Dec Jan Dec0
1
2
3
4
5
6
7
8Ever. Broad. Forest (TAPA, Brazil)
2002 2003month
LAI
24
Global Vegetation Workshop, University of Montana
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7Crops(RMSE=0.941, R 2=0.61)
GLOBCARBON
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7Crops(RMSE=1.23, R 2=0.35)
ECOCLIMAP
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7Crops(RMSE=1.45, R 2=0.307)
ECOCLIMAP
GLOBCARBON
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7Crops(RMSE=0.827, R 2=0.674)
CYCLOPES
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7Crops(RMSE=0.604, R 2=0.62)
CYCLOPES
GLOBCARBON
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7Crops(RMSE=1.32, R 2=0.485)
CYCLOPES
ECOCLIMAP
Scatter plot over the BELMANIP sites-class=crop
25
Global Vegetation Workshop, University of Montana
Scatter plot over the BELMANIP sites-class=ENF
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=1.27, R 2=0.527
GLOBCARBON
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=0.999, R 2=0.532
ECOCLIMAP
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=1.38, R 2=0.527
ECOCLIMAP
GLOBCARBON
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=1.3, R 2=0.587
CYCLOPES
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=1.27, R 2=0.553
CYCLOPES
GLOBCARBON
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=1.5, R 2=0.386
CYCLOPES
ECOCLIMAP
26
Global Vegetation Workshop, University of Montana
Scatter plot over the BELMANIP sites-class=EBF
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=2.09, R 2=0.021
GLOBCARBON
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=1.59, R 2=0.0389
ECOCLIMAP
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=2.44, R 2=0.0579
ECOCLIMAP
GLOBCARBON
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=1.74, R 2=0.486
CYCLOPES
MODIS
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=2.33, R 2=0.167
CYCLOPES
ECOCLIMAP
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7RMSE=1.48, R 2=0.254
CYCLOPES
GLOBCARBON
27
Global Vegetation Workshop, University of Montana
Direct validation: Product versus HR LAI Map
0 1 2 3 4 5 60
1
2
3
4
5
6
LAI map
MODIS
RMSE=1.4, R 2=0.566
0 1 2 3 4 5 6 70
1
2
3
4
5
6
7
LAI map
GLOBCARBON
RMSE=1.5, R 2=0.411
0 1 2 3 4 5 60
1
2
3
4
5
6
LAI map
ECOCLIMAP
RMSE=1.37, R 2=0.459
0 1 2 3 4 5 60
1
2
3
4
5
6
LAI map
CYCLOPES
RMSE=0.888, R 2=0.7
28
Global Vegetation Workshop, University of Montana
Results summary
++++++++++++++Seasonal
trends
++++++++++Magnitude
NA++++++++++In situ data
NA++++++++++++Spatial
gradient
NA++++++++++++global LAI
distribution
+++++++++++Missing data
CCRSCYCLGLOBCARMODISECOCLI
29
Global Vegetation Workshop, University of Montana
Pending issues
30
Global Vegetation Workshop, University of Montana
Concluding Remarks