1. GSICS Web Meeting, 17 November 2011
Intersatellite Calibration of Kalpana TIR & WV Channels with AIRS and IASI Hyperspectral Observations
Space Applications Centre
Indian Space Research Organisation
Ahmedabad (INDIA)
GSICS Web Meeting, 17 November 2011

2. Satellite Sensors and Data Used
Indian Geostationary Satellites
Kalpana : VHRR
INSAT-3A : VHRR
INSAT-3D : Imager (6 Channels – VIS, SWIR, MWIR, WV, TIR1, TIR2)
Sounder (18 IR + 1 Vis Channels)
Hyperspectral IR Sounders
IASI (Metop) 8461 IR-Channels (Morning/Evening Observations)
AIRS (Aqua) 2378 IR-Channel (Noon/mid-night Observations)
Intercalibration period
AIRS: 2010 (Jan, Apr, Jul, Oct)
IASI: (Jul-Oct)

3. Kalpana Launched in Sep 2002 Positioned at 74°E.
Full disk every 30 minutes
VHRR
VIS : m
WV : m
TIR : m
Spatial resolution (at nadir) :
2 km for VIS
8 km for WV and TIR channels

4. Aqua-AIRS and Metop-IASI spectral coverage with Kalpana SRFs for TIR and WV channels
AIRS: Atmospheric InfraRed Sounder
Polar Orbiting Aqua (2002)
Channels: (650 cm-1 to 2675 cm-1)
(3.74 μm μm)
Spectral resolution: /1200
Spatial Resolution: 13.5 Km at Nadir
Brightness Temperature
AIRS
IASI: Infrared Atmospheric Sounding Interferometer
Polar Orbiting Metop (2007)
Channels: (645 cm-1 to 2760 cm-1)
(3.62 μm μm)
Spectral resolution: cm-1 at SWIR
0.50 cm-1 at LWIR
(resampled at 0.25 cm-1)
Spatial Resolution: ~12 km at Nadir
IASI

5. Methodology
Convolved radiance of broadband sensor using ‘n’ number of hyperspectral sounders channels may be computed using:
Rconv is convolved broadband radiance, RAIRS is radiance of hyper-spectral sounder, superscript ‘i’ is hyper-spectral channel index,
SKalp is the sensor response function of Kalpana 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.
Missing or bad channels are interpolated from nearest 2 channels.
Conversion of radiances to brightness temperature by inverting the equivalent Planck’s function:
a1 (=1.052K) and a2 (=0.996) are the band correction coefficients and  (= cm-1) is central wavenumber for Kalpana TIR channel

6. Methodology……. Following steps are used:
Convolution of Kalpana-TIR radiance from AIRS radiances
environment uniformity test: StdDev of 3x3 AIRS and 5x5 Kalpana pixels (~50x50 km) < 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 | < 15 min)
Zenith angle collocation : | sec(AIRS) – sec(Kalp) | < 0.01
and AIRS & Kalp < 30°

7. Comparison of Kalpana Tb with AIRS Convolved Tb for Thermal IR Channel

10. Intercalibration of Kalpana IR and WV Channels with IASI
Data Source: IASI L1C received 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(AIRS) – sec(Kalp) | < 0.01
IASI & Kalpana < 10°
Spatial homogeneity test: Wang et al., JAOT 2009, vol.26,
(Std. Dev. of Kalpana Tb within one IASI IFOV) / (mean Tb of IASI FOR) < 0.01
Difference max(Tb) - min(Tb) within IASI IFOV <=3.0K

11. ISAI Observation geometry
IFOV
50 km
12 km

12. Kalpana IR/WV pixels within ISAI IFOV

13. Testing of the Intercalibration Procedure with Meteosat-7 TIR and WV Channels

14. Meteosat-7 vs IASI convolved
01 June 2011, 0230 GMT
EUMETSAT - GSICS webpage for Meteosat-7 vs IASI intercalibration

15. Intercalibration Results for Kalpana TIR and WV Channels (Jul-Oct, 2011)

16. Temporal collocation criteria : dt < 5 min

17. Kalpana TIR Channel Daytime (morning)
Temporal collocation < 30 minutes
Mean StDev = 1.3 K
(with AIRS: 0.41K day, 0.33K night
Clear sky only Tb > 285K)

18. Kalpana TIR Channel Night-time (Evening)
Temporal collocation < 30 minutes
Mean StDev = 1.3 K
(with AIRS: 0.41K day, 0.33K night
Clear sky only Tb > 285K)

19. Kalpana WV Channel Daytime (morning)
Temporal collocation < 30 minutes
Mean StDev = 0.92 K

20. Kalpana WV Channel Night-time (Evening)
Temporal collocation < 30 minutes
Mean StDev = 0.96 K

21. Conclusions
Kalpana TIR Channel intercalibrated with AIRS radiances for 2011 (clear sky conditions)
Std. Dev. of differences very small K
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

22. Thank You