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 Kumarkiran@sac.isro.gov.in

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