Recent Calibration and Validation Activity for Remote Sensing at NIST

1
Recent Calibration and Validation Activity for
Remote Sensing at NIST

• Dr. Raju Datla
• Optical Technology Division
• NIST Gaithersburg, MD 20899
• CEOS, WGCV 18
• ESA-ESRIN, Frascati, Italy
• June 5 - 7, 2001

2
Outline

• Goals of NIST Cal/Val Plans
• Activity During the Past 6 Months
• Scripps - NISTAR
• Scripps - EPIC
• MOBY
• Other support activities.
• Conclusion

3
NIST Cal/Val Plansfor Optical Remote Sensing
Programs

• Develop the needed facilities and instrumentation
• SIRCUS
• Involve from Inception
• Component and System End to End Testing
• Promote Well Conceived Calibration Plans from the
  Start
• Activities like Algorithm Theoretical Basis
  Document (ATBD) development for MODIS
• Participate in Cal/Val as needed in on-going
  Programs

4
SIRCUS Calibration Facility at NIST
A variety of tunable lasers
sphere (not shown) on translation stages
Laser output fiber-coupled into an integrating
sphere
wavemeter
pump laser beam
optical fiber
cw dye laser
ultrasonic bath (removes effects of speckle)

• Produces a spatially uniform, monochromatic,
  broadly tunable source of known radiance (0.1
  uncertainty uses transfer detectors and the NIST
  cryogenic radiometer)
• With ?? lt 0.001 nm, result is the true radiance
  (or irradiance) responsivity high flux levels
  give excellent signal to noise ratios optics of
  radiometer filled
• Accurate determination of in-band and
  out-of-band component

5
Triana Spacecraft
The Triana Mission involves four separate
instruments EPIC - 2kX2k CCD with 10 filter
channels NISTAR - Whole Disk Earth
Radiance Plas-Mag - Solar Physics and Space
Weather PHA - High Energy Particles
6
Instrument Requirements

• Three Radiometers
• A visible to far infrared (0.2 µm to 100 µm)
  channel to measure total radiant power.
• A solar (0.2 µm to 4 µm) channel to measure
  reflected solar IR and visible radiation.
• A near infrared (0.7 µm to 4 µm) channel to
  measure reflected IR solar radiation.
• A photodiode (0.2 µm to 1.1 µm) channel
• Monitoring of radiometer filter elements.
• Visible and Near IR reflected solar radiation at
  100mS sampling rate.

7
NISTAR Assembly Overview

• Actual Dimensions
• Instrument 33.8 (l) x 25.9 (w) x 58.4 (h) cm
• ICE 33.8 (l) x 25.9 (w) x 25.4 (h) cm
• Radiometer Assy 33.1 (h) x 24.1 (dia) cm
• Mass Budget 23.5 kg
• Radiometer Assy, 7 kg
• ICE Assy, 1 kg
• Cable Assy, 2kg
• Mechanical Interface to S/C
• Radiometer FOV - 1.0
• Radiometer FOR - 7.0
• Thermal Interface to S/C
• ICE 20 10 C
• Heat transfer rate 2-3 W conductive to S/C
• View Factors S/C bus at 20C
• Passed Vibration Test 14g
• Power Consumption 42.3 watts _at_ 30 VDC
• Command/Data Interface
• Mil Std 1553

Radiometer Assembly
Interface Control Electronics (ICE)
8
Characterization of Absolute Scale

• Power Mode _at_ 632 nm
• Comparison to silicon-photodiode trap detector
  transfer standard.
• Trap is calibrated with High Accuracy Cryogenic
  Radiometer (HACR)
• Measurements with instrument in cryovac chamber
  using a stabilized HeNe laser and brewster
  windows.
• Spectral Irradiance Mode
• Use integrating sphere and tunable laser sources.
• Provide improved filter transmission results.

9
Optical Calibration Setup for Power Mode
10
Irradiance Mode Calibration at the SIRCUS Facility
During calibration of NISTAR at the SIRCUS
facility, the instrument was in a thermal-vacuum
chamber to simulate the space environment. It
viewed the output of a laser- illuminated
integrating sphere (green), simulating the
geometry of the view of Earth from L1.
The integrating sphere was on a translation
stage, and the laser was fiber-optically fed.
This enabled the source-to-detector distance to
be varied.
11
Summary Scripps -NISTAR

• The goal for the standard uncertainty for
  absolute power measurements was 0.1
• The difference from what was expected based on
  component by component characterizations and
  systems testing was 1 to 2.
• Recognized that internal reflections and
  scattering are introducing more uncertainty at
  the systems level.
• Improvement Replace with Wedged filters and
  Retest.

12
Scripps-EPIC Overview

• Scripps-EPIC - Earth Polychromatic Imaging Camera
• Full-color image of entire sunlit side of Earth
  every 15 minutes
• On Triana spacecraft at L1
• Measure ozone, aerosols, clouds, vegetation
• Optical path
• 30 cm Cassegrain telescope
• Collimating optics
• 2 wheels with interference filters for 5 UV, 3
  VIS, 2 NIR channels
• Slotted wheel for variable exposure times
• 2048 x 2048 CCD array

13
Scipps-EPIC Radiometric Calibration

• Goals 3 absolute, 1 band-to-band, 0.1
  pixel-to-pixel uncertainties in spectral radiance
  responsivity
• Measurements at Lockheed-Martin in thermal-vacuum
  chamber through window in Dec. 2000
• Sources of radiance
• Laurel hemisphere 7 levels
• Xe arc / Spectralon plaque
• Detectors calibrate radiance of sources and
  monitor during operation
• Filter radiometers with filters at EPIC channels
• Spectroradiometer at UV wavelengths
• Traceable to NIST primary standards

14
Scripps-EPIC Participants
NISTPersonnel
EPIC
15
Scripps-EPIC Source Setup
Laurel Hemisphere
Filter Radiometer
EPIC
16
Scripps-EPIC Xe/Plaque
Plaque
Lamp Housing
Filter Radiometer
EPIC
17
Scripps-EPIC Results

• Motivation separately determine parameters in
  measurement equation L (S Sd) / (t R) ,
• L radiance, S signal (linearity), Sd dark
  signal, t exposure time, R responsivity
• Preliminary results from measurements at
  Lockheed-Martin
• Signal is slightly non-linear at largest values
• Dark signal is constant at operating temperatures
• Exposure time is a complicated function of
  direction of wheel movement, slot, and
  temperature
• Responsivity is approximately that predicted from
  calculations, and is sufficient for operation
• Future measurements at NASA-Goddard
• Flat-field CCD array using a collimated source

18
Summary Scripps -EPIC(Earth Polychromatic
Imaging Camera)

• Developed calibration and characterization
  methodology.
• NIST developed, characterized and calibrated two
  5 channel filter radiometers
• provided traceability and
• determined radiance of the integrating sphere and
  the plaque sources used to calibrate the EPIC
  camera.

19
Marine Optical Spectrograph (MOS) for Marine
Optical Buoy (MOBY)
MOS
Blue Spectrograph
Red Spectrograph
Sphere source for calibration Water-leaving
radiance from oceans
Dichroic Beamsplitter
Radiometric Calibrations Stability
Wavelength Calibrations Others ..
Temperature Effects Effects of Stray Light
Uncertainty Sources
20
The Marine Optical Spectrographic (MOS) system is
used in two configurations one for the Marine
Optical Buoy (MOBY) and as a shipboard profiler.
Both systems are used for vicarious calibration
of satellite ocean color sensors, e.g. MODIS,
SeaWiFS, OCTS, POLDER, and IRS1-MOS.
Band-averaged normalized water-leaving radiances,
LWN's are reported by the MOBY team,
corresponding to data sets from MOBY at the
Lanai, Hawaii site and various sites for the MOS
profiler. For MODIS and SeaWiFS, band-averaged
LWN's are required for the range 412 nm to 670
nm. Here we report on the characterization of
stray light in the MOS profiler system. For the
first time, a rigorous study was possible using a
broadly tunable laser facility. We report
preliminary results for correction factors that
are required to assess the effect of stray light
on the derived up-welling radiance, based on
characterizations at NIST of the MOS Profiler.
21
Motivation for Stray Light Work
Red Spectrograph

• Circled Region Lu(?) derived using the two
  spectrographs in MOBY or the MOS Profiler
  disagree in their region of overlap degree of
  discrepancy is depth-dependent
• But at 412 and 440 nm 5 agreement with
  independent filter radiometers
• Stray light was suspected (a typical issue with
  single grating spectrographs used with sources of
  different spectral shapes)
• NOAA and NIST addressed the problem using tools
  available at the time
• New facility at NIST provides rigorous solution

Blue Spectrograph
22
Stray Light in Spectrographs

• Spectrograph operation Spectral separation by
  optical interference of specular reflections from
  gratingmaps to CCD columns
• Scattering is present Not all of the energy is
  in the specular beam, there is a
  forward-scattered (haze) and isotropic (diffuse)
  component (plus scattered light from remaining
  optical elements)
• Out of Band Result is the spectral selection
  is not ideal (ideal would be a Delta function)
• Filter Radiometer Same effect, but only one
  band per detector
• Issue for all single grating instruments

23
Corrected Radiances
Lu(?) Lcal(?) ST(?) SS(?)u / ST(?)
SS(?)cal

• A second iterative procedure is used to determine
  the corrected water-leaving radiance from the
  measured count rates
• Tested using a filtered integrating sphere source

Radiances
Correction Factor
24
MOS Correction MOS Data Sets
MOS 202 profiler data from representative
measurements of blue and green water (October
1999 and March 2001) were corrected using the MOS
SIRCUS results. The correction does not include
any effects of the second order interreflections.
At 412 nm, the preliminary corrections to the
MOS upwelled Lus are between 3 and 6.
Correction factor two ocean color bands
Corrected radiances
25
Application to MOBY

• MOBY vs. MOS Profiler on SIRCUS
• MOBY stray light characterization must be done at
  Snug Harbor
• Two MOSs are used on MOBY (interchanged each
  deployment)
• The MOSs in MOBY are stable but unique, so the
  algorithm correction parameters will be different
• These MOSs can be studied on MOBY, where MOS is
  integrated with the fiber optic inputs, or as
  separate optical systems
• Required Measurements
• scans with tunable laser for bandwidth
• measurements with fixed lasers (e.g, 412, 458,
  476, 488, 514, 543, 612, and 633 nm) for out of
  band profile
• adequate characterization of 2nd order
  reflections
• validation using the absolute colored sphere
  source
• MOBY correction factors will be different from
  the MOS Profiler results presented here

26
Tests to Date at Snug Harbor
Diode laser
Near infrared tunable diode laser used
successfully with Ed and Lu port of MOS profiler
Monitor photodiode
MOS202 Ed
Initial tests with fixed wavelength lasers appear
promising
Sphere
Lu MID
Ar-Ion Laser
27
Issues, Plans, and Summary

• Tunable blue laser for fine scans of blue
  spectrograph at Snug Harborissue under study
• NIST deployment in July and September 2001 to
  execute MOBY characterizations
• Validation of stray light correction algorithm
  using colored source with MOBYs and MOSs
• In situ validation using SIRCUS-characterized
  ocean filter radiometers during a MOCE cruise
  (winter 2001/2002)
• Fully correctable issue in instruments of proven
  stability will result in ocean color data set of
  the highest possible accuracy

Sponsors NOAA/NESDIS, NIST, and NASA/GSFC
(MODIS Science Team, SeaWiFS Project, SIMBIOS
Project)
28
Other Activities at NIST

• For EOS SORCE, a solar irradiance mission of Gary
  Rottman et al.
• the active cavity was tested to measure the
  reflectance and found to be below the level as
  expected.
• the apertures for the instrument are currently
  being calibrated at NIST
• Six witness sample Infrared filters for GOES
  Imager and Sounder instruments were received from
  ITT and NIST measured their transmittance as a
  function of temperature and incidence angle to
  simulate the operating conditions of the
  instruments.
• Tentative comparison between the ITT data and the
  NIST data shows differences. Work is in progress
  at NOAA to understand the difference and its
  effect on measurements.
• NIST participated in the second infrared
  Radiometer calibration and inter-comparison
  workshop at Miami, 28 May - 1 June 2001.