long-term assessment and recalibration of fy-1/3 virr and...
TRANSCRIPT
Long-term assessment and recalibration of FY-1/3 VIRR and MERSI reflective solar bands
Ling SUN, Hong QIU, Lin CHEN, Liyang ZHANG, Di XIAN, Na XU, Xiuqing HU, Peng ZHANG
National Satellite Meteorological Center, China Meteorological Administration, Beijing, China
The Medium Resolution Spectral Imager (MERSI) and Visible Infrared Radiometer (VIRR) are two major multi-spectral
imaging instruments onboard the second generation polar orbiting meteorological satellite Fengyun-3 (FY-3) in China.
The VIRR is inherited from Multispectral Visible Infrared Scanning Radiometer (MVISR) of the first generation polar
orbiting meteorological satellite (FY-1). The Fengyun satellite historical dataset reprocessing project carried out the task
of FY-1/3 VIRR and MERSI reprocessing to generate the long time series data record. The on-orbit radiometric response
changes of FY-1C/D VIRR, FY-3A/B/C VIRR and FY-3A/B MERSI reflective solar bands are derived using stable target
tracking such as desert, ocean and cloud. The long-term radiometric quality is evaluated using multisite calibration
tracking (MST), deep convective cloud (DCC) and simultaneous nadir overpass (SNO) methods. The dataset is re-
calibrated mainly using the daily coefficients derived by MST. It shows that the long-term varying of sensor radiometric
response is corrected and the radiometric stability is improved.
INTRODUCTION
METHOD
DATA
Acknowledgement: The research is supported by National Key R&D Program of China (No. 2018YFB0504900, 2018YFB0504905).
• Evaluation:
1. Comparison with radiometric references: radiometric references : PICs (desert/ocean/DCC), reference
instrument(SNO).
2. Find instrumental characteristics.
• Re-calibration:
1. Correction for the response time variation: MST trending model.
2. Correction for other potential issues: stray light, nonlinearity,
temperature dependence.
3. Correction for the radiometric bias : MODISFY-3FY-1.
RESULTS
Band CW
(m) BW
(m) IFOV (m)
NEρor NET
Dynamic Range (ρ or T)
1 0.470 0.05 250 0.3% 100% 2 0.550 0.05 250 0.3% 100% 3 0.650 0.05 250 0.3% 100%
4 0.865 0.05 250 0.3% 100% 5 11.25 2.5 250 0.54K 330K 6 1.640 0.05 1000 0.08% 90% 7 2.130 0.05 1000 0.07% 90% 8 0.412 0.02 1000 0.1% 80% 9 0.443 0.02 1000 0.1% 80%
10 0.490 0.02 1000 0.05% 80% 11 0.520 0.02 1000 0.05% 80% 12 0.565 0.02 1000 0.05% 80% 13 0.650 0.02 1000 0.05% 80% 14 0.685 0.02 1000 0.05% 80% 15 0.765 0.02 1000 0.05% 80% 16 0.865 0.02 1000 0.05% 80% 17 0.905 0.02 1000 0.10% 90% 18 0.940 0.02 1000 0.10% 90% 19 0.980 0.02 1000 0.10% 90% 20 1.030 0.02 1000 0.10% 90%
CW: Central wavelengths; . BW: Bandwidths; IFOV: Instantaneous field of view (nadir).
1/ Ref RRPDif Mea
VIRR and MERSI are both cross-track scanning radiometer. FY-1A/B/C/D were launched on
Sept. 7, 1988, Sept. 3, 1990, May 10, 1999, and May 15, 2002, respectively. FY-3A/B/C were
launched on May 27, 2008, Nov. 5, 2010, and Sept. 23, 2013,respectively.
Table 2 FY-3A/B/C MERSI spectral specification.
• Sensor Normalized Response Sensor response was derived using MST and DCC methods. For VIRR, it showed faster attenuation at smaller wavelength and obvious seasonal variation at 865nm. For MERSI, it showed large degradation at short wavelength and 1030nm. For both VIRR and MERSI, there was larger seasonal change for FY-3B.
Figure 3 Normalized response of FY-3A/B/C VIRR using MST method
• Calibration Performance The dataset is re-calibrated using the daily coefficients derived by MST. The re-calibrated TOA reflectance was evaluated using SNO and stable targets. The lifetime stability and inter-platform consistency were improved.
Figure 6 PDif of operational and re-calibrated L1 TOA reflectance at desert targets
Figure 5 Normalized response of FY-1D VIRR, FY-3 VIRR & MERSI using MST method
Ban
d.
FY-1A/B
VIRR-1
FY-1C/D
VIRR-2
Dyna
mic
FY-3A/B/C
VIRR-3
Dynam
ic
1 0.58-0.68 0.58-0.68 90% 0.58-0.68 100%
2 0.725-1.10 0.84-0.89 90% 0.84-0.89 100%
3 0.43-0.53 3.55-3.95 3.55-3.95
4 0.53-0.58 10.3-11.3 10.3-11.3
5 10.5-12.5 11.5-12.5 11.5-12.5
6 1.56-1.64 80% 1.56-1.64 90%
7 0.43-0.48 50% 0.43-0.48 50%
8 0.48-0.53 50% 0.48-0.53 50%
9 0.53-0.58 50% 0.53-0.58 50%
10 0.900-0.960 90% 1.325-1.395 90%
Table 1 FY-1/3 VIRR spectral specification.
Figure 2 Time range of FY-1/3 dataset.
Figure 1 Fixed site positions of PICs.
7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1
0.6
0.8
1
2008 2009 2010 2011 2012 2013 2014
Date
No
rm
alized
R
esp
on
se
FY-3A VIRRX
7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1
0.6
0.8
1
1.2
2008 2009 2010 2011 2012 2013 2014
Date
No
rm
alized
R
esp
on
se
FY-3A VIRRX
B01
B02
B06
B07
B08
B09
B10
11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11
0.6
0.8
1
2014 2015 2016 2017 2018
Date
No
rm
alized
R
esp
on
se
FY-3C VIRRX
11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 110.7
0.8
0.9
1
1.1
2014 2015 2016 2017 2018
Date
No
rm
alized
R
esp
on
se
FY-3C VIRRX
B01
B02
B06
B07
B08
B09
B10
1213579111357 91113 5791113579111357911135791113579111357 9110.7
0.8
0.9
1
1.1
2010 2011 2012 2013 2014 2015 2016 2017 2018
Date
No
rmal
ized
Res
po
nse
FY-3B MERSI
1213579111357 91113 5791113579111357911135791113579111357 9110.6
0.7
0.8
0.9
1
1.1
20102011 2012 2013 2014 2015 2016 2017 2018
Date
No
rmal
ized
Res
po
nse
FY-3B MERSI
1213579111357 91113 5791113579111357911135791113579111357 9110.7
0.8
0.9
1
1.1
20102011 2012 2013 2014 2015 2016 2017 2018
Date
No
rmal
ized
Res
po
nse
FY-3B MERSI
1213579111357 91113 5791113579111357911135791113579111357 9110.7
0.8
0.9
1
1.1
20102011 2012 2013 2014 2015 2016 2017 2018
Date
No
rmal
ized
Res
po
nse
FY-3B MERSI
B03
B04
B13
B14
B15
B16
B01
B02
B08
B09
B10
B11
B12
B20
B06
B07
B17
B19
79111357 91113 57911135 79111357 91113 57911135 79110.7
0.8
0.9
1
1.1
2008 2009 2010 2011 2012 2013 2014 2015
Date
Nor
mal
ized
Res
pons
e FY-3A MERSI
79111357 91113 57911135 79111357 91113 57911135 79110.4
0.6
0.8
1
2008 2009 2010 2011 2012 2013 2014
Date
Nor
mal
ized
Res
pons
e
FY-3A MERSI
79111357 91113 57911135 79111357 91113 57911135 79110.7
0.8
0.9
1
1.1
2008 2009 2010 2011 2012 2013 2014
Date
Nor
mal
ized
Res
pons
e FY-3A MERSI
79111357 91113 57911135 79111357 91113 57911135 79110.7
0.8
0.9
1
1.1
2008 2009 2010 2011 2012 2013 2014
Date
Nor
mal
ized
Res
pons
e FY-3A MERSI
B03
B04
B13
B14
B15
B16
B01
B02
B08
B09
B10
B11
B12
B20
B06
B07
B17
B19
79111357 91113 57911135 79111357 91113 57911135 79110.7
0.8
0.9
1
1.1
2008 2009 2010 2011 2012 2013 2014 2015
Date
Nor
mal
ized
Res
pons
e FY-3A MERSI
79111357 91113 57911135 79111357 91113 57911135 79110.4
0.6
0.8
1
2008 2009 2010 2011 2012 2013 2014
Date
Nor
mal
ized
Res
pons
e
FY-3A MERSI
79111357 91113 57911135 79111357 91113 57911135 79110.7
0.8
0.9
1
1.1
2008 2009 2010 2011 2012 2013 2014
Date
Nor
mal
ized
Res
pons
e FY-3A MERSI
79111357 91113 57911135 79111357 91113 57911135 79110.7
0.8
0.9
1
1.1
2008 2009 2010 2011 2012 2013 2014
Date
Nor
mal
ized
Res
pons
e FY-3A MERSI
B03
B04
B13
B14
B15
B16
B01
B02
B08
B09
B10
B11
B12
B20
B06
B07
B17
B19
121 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11
0.6
0.8
1
2010 2011 2012 2013 2014 2015 2016 2017 2018
Date
No
rm
alized
R
esp
on
se
FY-3B VIRR
121 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11 1 3 5 7 9 11
0.6
0.8
1
1.2
2010 2011 2012 2013 2014 2015 2016 2017 2018
Date
No
rm
alized
R
esp
on
se
FY-3B VIRRX
B01
B02
B06
B07
B08
B09
B10
Figure 4 Normalized response of FY-3A/B MERSI using MST method
20002001200220032004200520062007200820092010201120122013201420152016 2017201820190.6
0.7
0.8
0.9
1
1.1
Date
No
rmal
ized
Res
po
nse
VIRRXMERSI VIRR B2 865nm MERSI B4 865nm
FY-1D VIRR
FY-3A VIRR
FY-3B VIRR
FY-3C VIRR
FY-3A MERSI
FY-3B MERSI
FY-3D MERSI
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018-0.4
-0.35-0.3
-0.25-0.2
-0.15-0.1
-0.050
0.050.1
0.150.2
Date
PD
if
VIRRMERSI VB2
FY-1C VIRR
FY-1D VIRR
FY-3A VIRR
FY-3B VIRR
FY-3C VIRR
FY-3A MERSI
FY-3B MERSI
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018-0.5
-0.45-0.4
-0.35-0.3
-0.25-0.2
-0.15-0.1
-0.050
0.050.1
0.150.2
Date
PD
if M
ean
VIRR B2 MERSI B4 865nm
FY-1C VIRR
FY-1D VIRR
FY-3A VIRR
FY-3B VIRR
FY-3C VIRR
FY-3A MERSI
FY-3B MERSI
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 20180
0.05
0.1
0.15
0.2
0.25
0.3
Date
PD
if S
td
VIRRMERSI VB2
FY-1C VIRR
FY-1D VIRR
FY-3A VIRR
FY-3B VIRR
FY-3C VIRR
FY-3A MERSI
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018-0.4
-0.35-0.3
-0.25-0.2
-0.15-0.1
-0.050
0.050.1
0.150.2
Date
PD
if
VIRRMERSI VB2
FY-1D VIRR
FY-3A VIRR
FY-3B VIRR
FY-3C VIRR
FY-3A MERSI
FY-3B MERSI
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018-0.4
-0.35-0.3
-0.25-0.2
-0.15-0.1
-0.050
0.050.1
0.150.2
Date
PD
if M
ean
VIRR B2 MERSI B4 865nm
FY-1D VIRR
FY-3A VIRR
FY-3B VIRR
FY-3C VIRR
FY-3A MERSI
FY-3B MERSI
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 20180
0.05
0.1
0.15
0.2
0.25
0.3
Date
PD
if S
td
VIRRMERSI VB2
FY-1D VIRR
FY-3A VIRR
FY-3B VIRR
FY-3C VIRR
FY-3A MERSI
FY-3B MERSI
Re-processed V1
Operational
0.63 μm Operational DCC Ref
0.63 μm Re-processed V1 DCC Ref
SUMMARY
Vicarious methods are used to evaluate the FY-1/3 VIRR &MERSI RSB data performance and correct the time variation of instrument response. The data performance could be improved after re-calibration simply using the MST trending result, while the absolute accuracy may be improved. The time factor in re-calibration model needs to be adjusted , especially for the early phase in orbit and blue bands. The seasonal features in normalized response and PDif time series reveal unresolved issues related to instrumental characteristics, such as stray light, nonlinearity, temperature dependence, etc. Result differences between MST, DCC , SNO need be investigated to obtain consistency.
Figure 7 TOA reflectance of operational and re-calibrated L1 at DCC targets
Figure 8 Lifetime TOA reflectance histogram, reflectance and PDif time series of re-calibrated L1 at SNO targets against Aqua/MODIS