radiometric calibration and atmospheric correction
TRANSCRIPT
David R. Thompson1
Bo-Cai Gao2
Robert O. Green1
1 Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA2 Naval Research Laboratory, Washington, DC
Copyright 2015 California Institute of Technology. All Rights Reserved.
US Government Support Acknowledged.
Radiometric calibration and
atmospheric correction
6/18/2015 [email protected] 1
6/18/2015 [email protected] 2
Raw DNs
Radiance (W/nm/sr/cm2)
Reflectance (HDRF)
Part 1: Radiometric Calibration
Part 2: Atmospheric Correction
NASA/JPL Portable Remote Imaging Spectrometer (PRISM)
A motivating example
from PRISM
6/18/2015 [email protected] 3
0.4 0.45 0.5 0.55 0.6 0.650
0.002
0.004
0.006
0.008
0.01
Rrs
0.4 0.45 0.5 0.55 0.6 0.65
−2
−1
0
1
2
x 10−3
Wavelength (microns)
Res
iduu
m
Wavelength (microns)
PRISM
flightline
M0 Buoy
retrieved
Pajaro River
retrieved
Wavelength (microns)
In situ
In situ
D. R. Thompson, F. Siedel, B.-C. Gao, M. Gierach, R. Kudela, R. O. Green, P. Mouroulis. Optimizing Solar
Irradiance for Coastal Spectroscopy. Geophysical Research Letters (2015, in press).
0.4 0.45 0.5 0.55 0.6 0.650
0.002
0.004
0.006
0.008
0.01
Rrs
0.4 0.45 0.5 0.55 0.6 0.65
−2
−1
0
1
2
x 10−3
Wavelength (microns)
Res
idu
um
In-situ data courtesy Raphe Kudela, UCSC
0.4 0.45 0.5 0.55 0.6 0.650
0.002
0.004
0.006
0.008
0.01
Rrs
0.4 0.45 0.5 0.55 0.6 0.65
−2
−1
0
1
2
x 10−3
Wavelength (microns)
Res
iduu
m
Two issues…
6/18/2015 [email protected] 4
UV falloffSpikes?
Wavelength (microns)
6/18/2015 [email protected] 6
Remove dark current levels
Remove electronic
effects
Apply radiometric calibration
Apply flat fieldFind
zero
Translate to Radiance
Raw DNs
Radiance (W/nm/sr/cm2)
NASA/JPL Portable Remote Imaging Spectrometer (PRISM)
Radiometric Calibration
Calibration challenges:
radiometry
6/18/2015 [email protected] 7
0 500 1000 1500 2000 25000
0.5
1
1.5
2
2.5
3x 10
−5
Wavelength (nm)
W/(
cm2 n
m)
Typical red-rich calibration source
Spectral
response
affects the
estimated
radiometry
6/18/2015 [email protected] 8
0.4 0.5 0.6 0.7 0.8 0.9 1
−4
−2
0
Panel scan: ang20140610t232914
PS
F (
Lo
g s
cale
)
0.4 0.5 0.6 0.7 0.8 0.9 10
5000
10000
DN
s
0.4 0.5 0.6 0.7 0.8 0.9 10.85
0.9
0.95
1
1.05
Wavelength (microns)
Ratio
Illuminated panel
measurement
Deconvolved
Deconvolved
Measured
Spectral response
having non-Gaussian
tails
6/18/2015 2JPL/Caltech PROPRIETARY — Not for Public Release or Redistribution. The technical data in this document is controlled under the U.S. Export Regulations, release to foreign persons may require an export authorization.
0.38 0.4 0.42 0.44 0.46 0.48 0.5 0.52 0.54 0.56 0.580
0.5
1
Wavelength (micron)
Val
ue
PRISM sampling
TNO2
TO3
TuT
d
s
ra
Irradiance 1250
2500
Irra
dia
nce
W/m
2 /mm
6/18/2015 [email protected] 11
Gas absorption terms
Path reflectance
Optimizing irradiance estimates
• Hypothesis: fine spectral sampling (~3nm) causes sensitivity to sampling of the solar irradiance (and intrinsic uncertainty)
• Solution: modify an irradiance estimate using a smooth in-scene reference(here, a concrete surface)
6/18/2015 [email protected] 12
Optimizing irradiance estimates
6/18/2015 [email protected] 13
Penalize difference vs. smoothed reflectance
Penalize large perturbations
LevenbergMarquardt
Agreement
with in situ Rrs
is improved
D. R. Thompson, F. Siedel, B.-C. Gao, M. Gierach, R. Kudela, R. O. Green, P. Mouroulis. Optimizing Solar Irradiance for Coastal Spectroscopy. Geophysical Research Letters (2015, in press).
6/18/2015 3JPL/Caltech PROPRIETARY — Not for Public Release or Redistribution. The technical data in this document is controlled under the U.S. Export Regulations, release to foreign persons may require an export authorization.
Wavelength (microns) Wavelength (microns)
M0 Buoy,
Rrs retrieval using different
solar irradiance spectra
Pajaro River
Aerosols are a persistent challenge
In-situ data courtesy Sherry Palacios and Liane Guild, NASA Ames; Raphe Kudela, UCSC
In situ
remote
Concluding thoughts
• Ocean observations place extreme
requirements on both calibration and
atmospheric correction
• Is there a common root cause to both
issues (far tails of the SRF)?
• Underscores need for spectral uniformity
6/18/2015 [email protected] 16
Thanks
• The PRISM team, including Zakos
Mouroulis, Byron Van Gorp, Mark
Helmlinger, Scott Nolte, Sarah Lundeen
• Felix Seidel, Heidi Dierssen, Michelle
Gierach, John Fontenla, Raphe Kudela,
6/18/2015 [email protected] 17