GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 1
Intersatellite Calibration of infrared sensors onboard Indian Geostationary Satellites using
LEO Hyperspectral Observations
A S Kiran [email protected]
P K Pal, C M Kishtawal, Pradeep Thapliyal, Munn ShuklaSpace Applications Centre
Indian Space Research OrganisationAhmedabad (INDIA)
Updates from GSICS members and ObserversIndian Space Research Organisation (ISRO)
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 2
INSAT-2E VHRR, CCD
Indian Meteorological Geostationary Satellites
1999
INSAT-1DVHRR
INSAT-2A/2BVHRR
1992/93
1990KALPANA-1
VHRR
2002
2013
INSAT-3DImager/ Sounder
INSAT-3A VHRR, CCD
2003
2015
INSAT-3DRImager/ Sounder
2016
GISATMX-LWIR/MX-VNIR/ Hys-VNIR/Hys-SWIR
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 3
INSAT-3A & Kalpana-1
Payload : (i) VHRR & CCD in INSAT 3A(ii) VHRR in Kalpana-1
• CCD Bands (µm)Vis (0.62 – 0.68), NIR (0.77 – 0.86), SWIR (1.55 – 1.69) Resolution : 1 km
• VHRR Bands (µm)– Visible : 0.55 – 0.75– Water vapour : 5.70 – 7.10 – Thermal Infra Red : 10.5 – 12.5
• Resolution (km) : 2 X 2 for Visible8 X 8 for TIR
Launch Date & Location : Kalpana-1 - 74.0ºE (Sep 2002)INSAT 3A - 93.5ºE (Apr 2003)
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 4
• Spectral Bands (µm)Short Wave Infra Red : Six bands Mid Wave Infra Red : Five BandsLong Wave Infra Red : Seven BandsVisible : One Band
• Resolution : 10 km
INSAT-3D
• Spectral Bands (µm) ResolutionVisible : 0.55 - 0.75 1 kmShort Wave Infra Red : 1.55 - 1.70 1 kmMid Wave Infra Red : 3.70 - 3.95 4 kmWater Vapour : 6.50 - 7.10 8 kmThermal Infra Red – 1 : 10.30 - 11.30 4 kmThermal Infra Red – 2 : 11.30 - 12.50 4 km
6 Channel IMAGER 19 Channel SOUNDER
Launch Date: 26 Aug 2013Location: 83E
200
220
240
260
280
300
3.5 4 4.5 5
Brig
htne
ss T
empe
ratu
re (
K)
200
220
240
260
280
300
5 6 7 8 9
Brig
htne
ss T
empe
ratu
re (
K)
200
220
240
260
280
300
9 10 11 12 13 14 15
Brig
htne
ss T
empe
ratu
re (
K)
9 8 7 6 5 4 3 2 1
18 17 16 15 14 13 12 11 10
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 5
OCEANSAT–1 (1999)
MSMR, OCM
Megha-Tropiques (2011)
MADRAS, SAPHIR,ScaRaB, ROSA
Altimeter
SARAL (2011)
Indian LEO Weather SatellitesOCEANSAT–2
(2009)
OSCAT,OCM, ROSA
OCEANSAT–3 (2014)
OSCAT, OCM, ROSA
GCOM-W2
Scatterometer,AMSR-2
Curr
ent
Futu
re
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 6
Satellite Sensors and Data Used
Indian Geostationary Satelliteswww.mosdac.gov.in
Hyperspectral IR SoundersIASI (Metop-A/B) 8461 IR-Channels (Morning/Evening Observations)
Data available through EUMETCAST
AIRS (Aqua) 2378 IR-Channel (Noon/mid-night Observations)Data available through http://mirador.gsfc.nasa.gov/
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 7
Intercalibration of Kalpana
Nor
mal
ized
SR
F
0
1
Nor
mal
ized
SR
F
0
1
Nor
mal
ized
SR
F
0
1
Nor
mal
ized
SR
F
0
1
IASI
AIRS
Brig
htne
ss T
empe
ratu
re
AIRS: Atmospheric InfraRed SounderPolar Orbiting Aqua (2002)
Channels: 2378 (650 cm-1 to 2675 cm-1)(3.74 μm - 15.4 μm)
Spectral resolution: ν/∆ν≈1200
Spatial Resolution: 13.5 Km at Nadir
IASI: Infrared Atmospheric SoundingInterferometer
Polar Orbiting Metop (2007)Channels: 8461 (645 cm-1 to 2760 cm-1)
(3.62 μm - 15.5 μm) Spectral resolution: 0.35 cm-1 at SWIR
0.50 cm-1 at LWIR(resampled at 0.25 cm-1)
Spatial Resolution: ~12 km at Nadir
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 8
(a) Day time observations
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1 6 11 16 21 30 6 12 17 22 27 2 7 12 17 22 27 1 6 11 22 27
Date (Month)
RM
SD /
BIA
S (K
)
StDevBIAS
April July OctoberJanuary
(b) Night time Observations
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1 7 13 18 25 3 8 13 19 24 29 5 12 18 23 28 3 8 13 20 26
Date (Month)
RM
SD /
BIA
S (K
)
StDevBIAS
January April July October
Midnight sun problem ??
Day-time Night-time Month RMSE BIAS SD N RMSE BIAS SD N
Jan 0.57 0.43 0.37 4641 1.88 1.84 0.38 4027 Apr 0.76 0.64 0.40 8578 1.52 1.50 0.28 6482 Jul 0.99 0.88 0.44 3342 1.58 1.54 0.36 2291 Oct 0.64 0.52 0.37 3908 1.52 1.49 0.29 4106
Total 0.75 0.63 0.41 20469 1.62 1.58 0.33 16906
Following steps are used:–Missing or bad channels are interpolated
from nearest 2 channels–Convolution of Kalpana-TIR radiance from
AIRS radiances–environment uniformity test:
SD of 3x3 AIRS and 5x5 Kalpana pixels < 2K–Land/Ocean flag: use only ocean pixels–Clear scene: Tb > 285K–Spatial Collocation:
< 0.2 (i.e. within one pixel resolution)–Temporal collocation: | tAIRS - tKalp | < 15m–Zenith angle collocation :
| sec(βAIRS) – sec(βKalp) | < 0.01 and βAIRS & βKalp < 30
Kalpana / AIRS Intercalibration
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 9
Kalpana / IASI Intercalibration
Data Source: IASI L1C through Eumetcast at SAC/ISRO in BUFR format
Methodology• Land/sea mask: Use only ocean pixels
• Temporal Collocation: 5, 15, 30 Minutes
• Spatial Collocation: – Within IASI IFOV (2x2 pixels)
• Zenith angle collocation:
| sec(βAIRS) – sec(βKalp) | < 0.01, and
βIASI & βKalpana < 10
• Spatial homogeneity test: Wang et al., JAOT 2009, vol.26,1843-1855
(SD of Kalpana Tb within one IASI IFOV) / (mean Tb of IASI FOR) < 0.01
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 10
Daily (Daytime) BIAS/ Standard Deviation of Differences (IR) Time Diff: 30 min
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1-Ju
l
11-J
ul
21-J
ul
31-J
ul
10-A
ug
20-A
ug
30-A
ug
9-S
ep
19-S
ep
29-S
ep
9-O
ct
19-O
ct
29-O
ct
Date (2011)B
IAS
/ S
tdD
ev (K
)
BIASStdDev
(Morning) Mean StDev = 1.3 K
Daily (Nighttime) BIAS/ Standard Deviation of Differences (IR) Time Diff: 30 min
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1-Ju
l
11-J
ul
21-J
ul
31-J
ul
10-A
ug
20-A
ug
30-A
ug
9-S
ep
19-S
ep
29-S
ep
9-O
ct
19-O
ct
29-O
ct
Date (July 2011)
BIA
S /
Std
Dev
(K)
BIASStdDev
(Evening) Mean StDev = 1.3 K
TIR Channel Kalpana/IASI (Jul-Oct, 2011)
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 11
Daily (Daytime) BIAS/ Standard deviation of difference (WV) Time Diff: 30 min
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1-Ju
l
6-Ju
l
11-J
ul
16-J
ul
21-J
ul
26-J
ul
31-J
ul
5-A
ug
10-A
ug
15-A
ug
20-A
ug
25-A
ug
30-A
ug
4-S
ep
9-S
ep
14-S
ep
19-S
ep
24-S
ep
29-S
ep
4-O
ct
9-O
ct
14-O
ct
19-O
ct
24-O
ct
29-O
ct
Date (2011)B
IAS
/ S
tdD
ev (K
)
biasStdDev
Daily (Nighttime) BIAS/ Standard deviation of difference (WV) Time Diff: 30 min
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1-Ju
l
6-Ju
l
11-J
ul
16-J
ul
21-J
ul
26-J
ul
31-J
ul
5-A
ug
10-A
ug
15-A
ug
20-A
ug
25-A
ug
30-A
ug
4-S
ep
9-S
ep
14-S
ep
19-S
ep
24-S
ep
29-S
ep
4-O
ct
9-O
ct
14-O
ct
19-O
ct
24-O
ct
29-O
ct
Date (2011)
BIA
S /
Std
Dev
(K)
BIASStdDev
(Morning) Mean StDev = 0.9K
(Evening) Mean StDev = 1.0 K
WV Channel Kalpana/IASI (Jul-Oct, 2011)
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 12
Summary of Kalpana Intercalibration
• Kalpana TIR Channel intercalibrated with AIRS radiances for 2011 (clear sky conditions)– Std. Dev. of differences very small 0.3-0.4K– Bias smaller during daytime with seasonal variations– Bias higher during nightime
• Kalpana TIR and WV channel intercalibration with IASI radiances (all-sky conditions)– Std. Dev. of differences high (1.3K TIR and 0.9K WV)– TIR: bias smaller during daytime with seasonal variations during– TIR: higher bias during nighttime– WV: lower bias during daytime and higher bias during nighttime
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 13
Data Source: IASI L1C (Eumetcast)Temporal Collocation: < 15 MinutesSpatial Collocation: within IASI pixel (12 km)
Zenith angle collocation:
maxzen = 0.1-0.4 depending upon the absorption
Spatial homogeneity test: Std. Dev. of INSAT-3D (7x7 pixel) and IASI (5x5) radiances within environment surrounding the target pixel
• Convolved radiance of broadband sensor using ‘n’ number of hyperspectral sounders channels may be computed using:
• Rconv is convolved broadband radiance, RIASI is radiance of hyper-spectral sounder, superscript ‘i’ is hyper-spectral channel index, SINSAT is the sensor response function of INSAT-3D channels at the central wavenumber of hyper-spectral channel ‘i’, and ‘n’ is the total number of hyper-spectral channels in broadband sensor’s SRF range.
( )( ) zen
zenleozengeo max_1
_cos_cos
<−
]Δ[/]Δ[11
νSνSRRn
i
iINSAT
n
i
iINSAT
iIASIconv ∑∑
==
=
200
220
240
260
280
300
3.5 4 4.5 5
Brig
htne
ss T
empe
ratu
re (
K)
200
220
240
260
280
300
5 6 7 8 9
Brig
htne
ss T
empe
ratu
re (
K)
200
220
240
260
280
300
9 10 11 12 13 14 15
Brig
htne
ss T
empe
ratu
re (
K)
9 8 7 6 5 4 3 2 1
18 17 16 15 14 13 12 11 10
INSAT-3D / IASI Intercalibration
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 14
14-Aug-201303:30Z
INSAT-3D Sounder / IASI Collocation map for GSICS
10-Aug-201303:31Z
10-Aug-201303:31Z
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 15
Night (13-16Z), Temporal Collocation : 15 Min, Spatial homogeneity: ENV_TB_SD = 2K
Oct-Nov 2013
Ch# BIAS(K)
SD(K)
MIR -0.1 1.1
WV -0.1 1.0
TIR1 -0.1 1.1
TIR2 -0.1 1.0
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 16
Night (13-15Z), Temporal Collocation : 5 Min, Spatial homogeneity: ENV_TB_SD = 1K
Oct-Nov 2013
Ch# BIAS(K)
SD(K)
MIR -0.1 0.7
WV -0.2 0.7
TIR1 -0.2 0.7
TIR2 -0.2 0.7
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 17
Day (3-6 Z)Temporal Collocation : 15 Min,
ENV_TB_SD = 3K
Oct-Nov 2013
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 18
Day (3-6 Z)Temporal Collocation : 5 Min,
ENV_TB_SD = 2K
Oct-Nov 2013
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 19
Time series of INSAT-3D Tb BIAS wrt Metop/IASI (Jan-Feb 2014)
-1.0
-0.5
0.0
0.5
1.0
3 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 29 30 31 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 21 22 23 24 25
Jan 2014 (Date) Feb-2014
BIA
S (K
)
CH#1CH#2CH#3CH#4CH#5CH#6CH#7CH#8CH#9CH#10CH#11CH#12CH#13CH#14CH#15CH#16
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 20
Time series of INSAT-3D Tb RMSD wrt Metop/IASI (Jan-Feb 2014)
0.0
0.5
1.0
1.5
2.0
3 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 29 30 31 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 21 22 23 24 25
Jan 2014 (Date) Feb-2014
RM
SD (K
)
CH#1CH#2CH#3CH#4CH#5CH#6CH#7CH#8CH#9CH#10CH#11CH#12CH#13CH#14CH#15CH#16
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 21
Status• Draft GSICS intercalibration ATBD is ready and
presented in GRWG meeting in EUMETSAT during 23-27 March 2014
• A test run is being conducted offline for near real-time GSICS intercalibration of INSAT-3D Imager/Sounder since March 2014.
• GSICS correction coefficients based on GSICS guidelines are being generated
• Past data of Kalpana is being processed
• Website for GSICS will be launched shortly
• Intercalibration of solar channel to be taken in near future
Thank You
GSICS-EP-15, 16-17 May 2014 (Guangzhou, China) 22