daily monitoring of vegetation conditions and ... · landsat 30m on 5/29/2016 . frequent but coarse...
TRANSCRIPT
Daily Monitoring of Vegetation Conditions and Evapotranspiration at Field Scale by
Fusing Multi-satellite Images
Feng Gao and Martha Anderson
USDA-Agricultural Research Service Hydrology and Remote Sensing Laboratory Beltsville, MD 20705 [email protected]
USDA is an equal opportunity provider and employer
MultiTemp 2017, June 27-29, 2017, Bruges, Belgium
May 6, 2016 May 11, 2016
May 27, 2016 June 11, 2016
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
Landsat 30m on 5/29/2016 MODIS 250m on 5/29/2016
frequent but coarse fine but infrequent
Sentinel-2A 10m on 6/26/2016
Landsat 30m on 8/17/2016
MODIS 250m on 6/10/2016 MODIS 250m on 6/26/2016
MODIS 250m on 7/23/2016 MODIS 250m on 8/17/2016
Combine Multi-sensor Data - High Temporal and Spatial Resolution Datacubes
Multi-sensor VI for Daily Monitoring of Vegetation Conditions
org. Landsat and S2
Harmonization
Data Fusion
Phenology Program Yield/Biomass
Variability
SR /VI Datacubes
Anomaly Analysis
Phenology Crop Growth
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
(i) Red
(j) NIR
(a) 4/18, ASTER (b) 4/26, AWiFS (c) 6/5, ASTER (d) 6/13, TM
(e) 7/7, AWiFS (f ) 7/23, ETM+ (g) 7/31, TM (h) 8/24, AWiFS
Multi-sensor Data Harmonization
Reference (MODIS NBAR)
Landsat 5/29/2016 MODIS 5/29/2016
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
Spatial and Temporal Adaptive Reflectance Fusion Model
STA
RFM
(3
0 m
)
STARFM
MO
DIS
La
ndsa
t
6/4/01 (155) 7/4/01 (185)
Reconstructed images
Observed
pairs
5/24/01 (144) 7/11/01 (192)
https://www.ars.usda.gov/research/software/download/?softwareid=432 (Gao et al., TRGS 2006)
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
N
Central Iowa (a rain-fed agricultural area) V
I and
ET
for
Dai
ly M
onit
orin
g of
Ve
geta
tion
Con
diti
on
MODIS NDVI (250m)
Jun 6 (157) Jul 4 (185) Aug 9 (221)
Sep 5 (248) Oct 1 (274) CDL (2011)
0.0 0.5 1.0 Yellow: corn; Green: soybeans
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
Jun 6 (157) Jul 4 (185) Aug 9 (221)
Sep 5 (248) Oct 1 (274) CDL (2011)
0.0 0.5 1.0 NDVI from fused data (30m) Yellow: corn; Green: soybeans
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
0.0
0.
5
1.0
Smoothed and Gap-filled Daily NDVI (30m), Apr. 1 – Nov. 1, 2011
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
2001 2006 2011 2014
Apr
May
Ju
n
Greenup
CD
L
Gao etc. (2017). Mapping crop progress at field scales using Landsat and MODIS imagery. Remote Sensing of Environment. 188:9-25.
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
Sep
Oct
N
ov
Dorm
ancy
Apr May Jun
Remote Sensing Phenology to
Crop Growth Stages
Gre
enup
dat
e R
elat
ions
hip
from
pre
viou
s ye
ars
Backcast Emerged Date (2014) VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
Yield Variability at Field Scale (corn, central Iowa, 2010)
0.7 0.8 0.9 1.0 max_NDVI
60 100 140 180 220 (Bu/Acre)
Monroe
Marshall
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
Central Plains Experimental Range (CPER) V
I and
ET
for
Dai
ly M
onit
orin
g of
Ve
geta
tion
Con
diti
on
Gap-filled and smoothed daily NDVI(30m) using original Landsat and the fused Landsat-MODIS data Central Plains Experimental Range (CPER) LTAR site, 2013
Data Sources: Landsat -8 125, 157,173, 221,237,269, 365 MODIS Daily (250m) 16-day BRDF Approach: -STARFM -Modified SG filtering
0.0
0.
2
0.4
0.
6
0.8
1.
0
J F M A M J J A S O N D
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
0.0
0
.5
1.0
Precipitation (May 15-Oct 1)
NDVI (August 1)
ANHP (g/m2)
MO
DIS
_ND
VI
ANHP: Aboveground Net Herbaceous Production
ANHP (g/m2)
LM_N
DV
I
… based on land surface temperature MULTI-SCALE SATELLITE ET RETRIEVAL
EVAPOTRANSPIRATION = EVAPORATION (non-beneficial use) + TRANSPIRATION (beneficial use)
Water use accounting
Irrigation efficiency
Crop stress and yield prediction
Crop water productivity (crop per drop)
0.0
0.2
0.4
0.6
0.8
1.0
4/1 5/1 5/31 6/30 7/30 8/29 9/28 10/28
ND
VI
Date
planted emerged silked dough dented mature harvestedStress (ET/PET)
[ ]∑ ××= HIETTLUEfPARPARYield iii *),,( maxεPeak
Dormancy
Clear Landsat
Landsat/ MODIS fusion
Remote sensing stages Phenological
stages
timeseries fit
Green-up
DD linkage VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
DATA FUSION: daily ET at field scale
Pola
r (L
ands
at)
Airb
orne
(U
SU a
ircra
ft)
Watershed
(60m)
Field scale (30m
)
SURFACE TEMPERATURE EVAPOTRANSPIRATION
GEO
(G
OES
Imag
er) Regional
(5km)
GEO
(G
OES
Sou
nder
) Continental (10km
)
Tem
pera
ture
(C)
Pola
r (M
ODI
S) Basin
(1km)
Latent Heat (Wm
-2) G
EO
(ISCC
P) G
lobal (25km
)
1 July 2002 – 10:30AM LST
corn soy
Hourly
Daily
1 LS – 16 day 2 LS – 8 day
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
GOES/MODIS/Landsat FUSION
GO
ES
(A
LEXI
) M
OD
IS
(Dis
ALEX
I) LA
ND
SA
T
(Dis
ALEX
I)
DOY 328 329 330 331 332 333 334 335 336
Landsat 5 Landsat 7
Spatial Temporal Adaptive Reflectance Fusion Model (STARFM)
Daily Evapotranspiration – Orlando, FL, 2002
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
0
2
4
6
8
10
150 160 170 180 190 200 210 220 230 240
Daily
ET (
mm
per
day
)
Day of Year
SMEX02 (Iowa) Soil Moisture Experiment 2002 Ames, Iowa Rainfed corn and soybean
4 km
SMEX02
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
0
2
4
6
8
10
150 160 170 180 190 200 210 220 230 240
Daily
ET (
mm
per
day
)
Day of Year
Rainfed soybean – SMEX02 (Iowa)
Reference ET Observed ET Landsat retrievals Landsat-only
RAIN
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
0
2
4
6
8
10
150 160 170 180 190 200 210 220 230 240
Daily
ET (
mm
per
day
)
Day of Year
Rainfed soybean – SMEX02 (Iowa)
Reference ET Observed ET Landsat retrievals Landsat-only Landsat-MODIS fusion
RAIN
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
0
50
100
150
200
250
300
150 160 170 180 190 200 210 220 230 240
Cum
ulat
ive
ET (m
m)
Day of Year
0
2
4
6
8
10
150 160 170 180 190 200 210 220 230 240
Reference ET Observed ET Landsat retrievals Landsat-only Landsat-MODIS fusion
Daily
ET (
mm
per
day
)
RAIN
Rainfed soybean – SMEX02 (Iowa) V
I and
ET
for
Dai
ly M
onit
orin
g of
Ve
geta
tion
Con
diti
on
Gallo Vineyards, Lodi CA
GRAPEX (Grape Remote sensing Atmospheric Profiling & Evapotranspiration eXperiment), 2013-2017
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
Landsat-7 and 8 (p44r33 and p43r33)
0 1 2 3 4
1/1 4/1 7/1 10/1 12/31
Landsat LAI VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
Landsat (30m)
Cumulative ET Cumulative LAI Max LAI
2013 Yield V
I and
ET
for
Dai
ly M
onit
orin
g of
Ve
geta
tion
Con
diti
on
Gallo Vineyards, Lodi CA
2015 Cumulative ET (mm) Cumulative ET (mm) 2016
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
NOAA – NESDIS – STAR GOES Evapotranspiration and Drought Product System (GET-D)
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
NOAA – NESDIS – STAR GOES Evapotranspiration and Drought Product System (GET-D)
Landsat (30m) GOES (4km)
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion
0
20
40
60
80
100
120
140
160
180
2000
2
4
6
8
10
0 50 100 150 200 250 300 350
ET (m
m/d
ay)
Day of year
Choptank (2014)
Rainfall (m
m/day)
ALEXI ET (4km) Observed ET Landsat retrieval Landsat-MODIS fusion Precipitation
4km
Sun, L., Anderson, M. C., Gao, F., Hain, C., Alfieri, J. G., Sharifi, A., McCarty, G. W., Yang, Y., Yang, Y., Kustas, W. P. and McKee, L. (2017), Investigating water use over the Choptank River Watershed using a multisatellite data fusion approach. Water Resour. Res.. doi:10.1002/2017WR020700. V
I and
ET
for
Dai
ly M
onit
orin
g of
Ve
geta
tion
Con
diti
on
Daily VI and ET Datacubes Improved monitoring of vegetation conditions Improved accounting of current water use and crop
water productivity (crop per drop) Monitoring changes in vegetation phenology and
water use with changing climate, land-use and human activities
Improved hydrologic monitoring (flood, drought, runoff) to better cope with extremes
Improved crop growth stages and stress detection for yield estimation
VI a
nd E
T fo
r D
aily
Mon
itor
ing
of
Vege
tati
on C
ondi
tion