Intercalibration between MTSAT1R and AIRS

1
Inter-calibration between MTSAT-1R and AIRS
Inter-calibration between MTSAT-1R and AIRS
MTSAT-1R Aqua AIRS
Aqua AIRS 2378 channels
Inter-calibration between MTSAT-1R and AIRS using
Taharas method
20/03/2007 World Weather Building Japan
Meteorological Agency Koji Kato
2
Self Introduction

• I am from Japan Meteorological Agency.
• I stay NESDIS from 26/2/2007 to 23/03/2007.
• (One Month)
• I was in charge of the Typhoon monitoring system
  for 5 years.
• I just started the inter-calibration work.
• Id appreciate it if you could speak slowly!

Hello! Everybody! Thank you for coming to my
presentation!
KOJI
3
Todays Topics

• What is the inter-calibration?
• How to do the inter-calibration?
• How to estimate the current spectral response
  function of geostationary satellites?

4
What is the inter-calibration

• Estimating the biases between satellites
• Needed for Multi-sensor Climate Data Records
• Blending environmental products requires
  minimizing systematic biases from different
  sensors
• Needed for Multi-sensor Climate Data Records and
  NWP

Quotation from GSICS meeting ppt. (M. Goldberg)
5
How to do the inter-calibration?

• Estimates the radiances of geostationary
  satellites using
• Hyper Spectral Infrared Sounder.
• Compare the radiances between geostationary
  satellites and HSIS

6
How to do the inter-calibration?

• Introduce the case between MTSAT-1R and AIRS
• Using Pre-Launch SRF
• ( Spectral Response Function)

7
MTSAT-1R
Provided by SS/L
8
AIRS

• Equipped on Aqua

9
How to estimate?

• We dont know how the SRF is now!

Im fine thank you! My SRF is same as Pre-Launch!
Hi! MTSAT! How is your SRF today?
Impossible! Cannot fly away to the space!
10
Taharas Virtual channel method

• Make the Virtual Channel
• using the SRF of geostationary satellites
• and AIRS

11
Taharas Virtual channel method

• Solve equations to get optimized wi.
• Please visit GSICS Page Taharas PPT

GEO SRF
Add AIRS SRFs like this!
This is Virtual Channel.
AIRS SRF
12
Generation of Virtual Channel from HSRS
Radiance observed by a broadband channel is
Radiance of a virtual channel (linear combination
of HSRS radiances) is
Find the optimized wi which makes J mininum
using Lagrange formula!
J is sum of the difference between Original SRF
and Virtual SRF
They should be approximately equal for any I(v),
then
To obtain wi, solve
where
Quoted from Taharas GSICS ppt.
13
AIRS Virtual Channel for MTSAT-1R IR2
Spectral Response Functions
MTSAT-1R IR2
AIRS virtual ch.
wi weights of AIRS virtual channel
Quoted from Taharas GSICS ppt.
14
Radiance and TB of Virtual Channel
Radiance of the virtual channel is obtained by
combining HSRS radiances
or
Inverse sensor Planck function is used to compute
brightness temperature of the virtual channel
where
Quoted from Taharas GSICS ppt.
15
Comparison of Simulated Radiance and TB

• Comparisons of simulated radiances and TBs show
  that the virtual channel eliminate the SRF
  differences

Simulated by LBLRTM (ver. 9.4) with HITRAN2000
and AER updates regarding the U.S. standard
atmosphere
Simulated radiances and TBs
Quoted from Taharas GSICS ppt.
16
Weak Spot!

• AIRS gaps occurs Gibbs Phenomenon!
• High Frequency!

IR1
IR3
17
Collocate data

• Collocate AIRS and MTSAT-1R (IR2) data
  (21/02/2007)
• Collocation conditions
• 15 min difference
• Uniformity check
• Cosecant of satellite zenith angles are in 5
  difference.

18
Time condition
Come on!

• Within 15 min
• Example
• 1000 and 1015 OK
• 1000 and 1016 NO

MTSAT-1R
19
Uniformity Check
MTSAT-1R

• Range 5 5 pixels
• T_max T_min lt 3 K

Temperature distribution is flat!
5
Target Pixel
5
20
Zenith Angle Check

• Within 5 Difference ( Secant1/Cos )
• 1/cos(?_g)-1/cos(?_a) ? 0.05 1/cos(?_g)

Satellite zenith angle
MTSAT-1R
?_g
Optical depth is almost same!
AIRS
?_a
21
Comparison Result1

• MTSAT-1R Temperatures are much higher

22
Why?

• Temperatures of MTSAT-1R are much higher than
  AIRS estimations!
• AIRS radiances are much less than MTSAT,
• Because

23
Dead channels!

• 249 of 2378 channels are dead!

24
Compensate dead channels

• Compensate the radiances of the dead channels
  using Average of 6 standard spectrums.
• kCARTA
• U.S. Standard
• Tropical
• Mid-summer winter
• Sub-arctic-summer winter

Tobin gave me the spectrum data! Thank you!
25
Spectrums
26
Comparison result 2

• Better than former case but .

Why so bad?
27
What happen?

• Pick up MAX, MIN Difference cases

28
Spectrums of two cases
29
Highest case

• Compensation is insufficient

Ave Six (Compensation)
30
Lowest case

• Compensation is excessive

Ave Six (Compensation)
31
How to adjust them?

• Adjust the ratio between AIRS spectrum and
  compensation spectrum!

32
Calculate the ratio between AIRS radiance and
Compensation radiance

• AIRS Radiance
• Compensation Radiance

Ratio
33
Ratio between radiances

• Ratio is Almost Constant in the IR2 band

Insufficient case
Constant...
Excessive case
34
Adjusted spectrum

• Good Approximation!

Insufficient case
Excessive case
35
Adjust the ratio for each FOV

• Good Result! Good Compensation!

36
Further Study

• Change the ratio more flexibly.
• Using RTMs (LBLRTM, CRTM, kCARTA,,,)
• to make more accurate compensation.
• Apply this method to the other channels.
• Improve Taharas method
• Improve the compensation method.
• (especially for water vapor channels.)

37
Cannot apply to WV channels
Ratio looks like Random!
Ozon, CO2,,,, channels are difficult too!
IR1
IR2
WV
38
How to estimate the Current SRF?
39
How to estimate the Current SRF?

• Estimate the current SRF of geostationary
  satellites using AIRS radiance.
• Inverse Taharas Method!

Taharas method
SRF
Radiance
40
Estimation Method 1

• RMTSAT MTSAT Radiance
• RAIRSi AIRS i th Channel Radiance
• wi Weight

Sorry! Now I dont go into details of the formulas
41
Estimation Result
Sensor anomaly? Data Error? Crosstalk? Compensa
tion Error?

• To be continued!

Sorry! Ill invertigate it later
What happen?
42
Further Study

• Investigate what happened!
• Using more accurate data
• (compensated by RTMs).

43
Thank you very much for your attention!
Any information is welcome! CRTM, McIDAS,
kCARTA, CrossTalk, Reference
Thank you very much For your attention!
44
Fin