The eighth EOS Calibration Panel Meeting was held at NASA/Goddard Space Flight Center (GSFC) on July 9, 10, and 11, 1996. Attendees included science and engineering representatives from the EOS AM-1, PM-1, and Chem platform instruments and from Landsat-7 Enhanced Thematic Mapper Plus (ETM+), the Active Cavity Radiometer Irradiance Monitor (ACRIM), the Tropospheric Emission Spectrometer (TES), the Stratospheric Aerosol and Gas Experiment III (SAGE III), and the Solar Stellar Irradiance Comparison Experiment (SOLSTICE). In addition, representatives from the National Institute of Standards and Technology (NIST), NASA Headquarters, the EOS Data and Information System (EOSDIS), NASA/GSFC Codes 400 and 700, the Canada Centre for Remote Sensing (CCRS), the University of Arizona Optical Sciences Center, and Northern Arizona University (NAU) attended the meeting.
Meeting Day 1: July 9, 1996
On the first day of the meeting, the EOS Calibration Scientist, Jim Butler, quickly reviewed the agenda for the three-day meeting. The first presentations were calibration status reports from each of the EOS instruments in attendance. Each presenter was instructed to summarize the calibration status of his/her instrument, clearly indicating any calibration-related problems.
Phil Slater of the University of Arizona Optical Sciences Center presented the calibration status report for the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) on behalf of Kohei Arai. The testing and calibration of the ASTER visible/near infrared (VNIR) and shortwave infrared (SWIR) instruments are reported to be progressing on schedule, but the thermal infrared (TIR) instrument is experiencing a delay due to scan mirror problems. The out-of-band rejection of the ASTER filters has been measured but a time line for the review of these data by the science team has not been established. Slater also stated that he was not sure if radiation exposure tests have been performed on the ASTER instrument. An optical defect in an ASTER VNIR filter was determined to affect the calibration of 3 ASTER pixels. This effect was discovered when illuminating the ASTER instrument with an F/84 beam but was not apparent when illuminating the instrument with an F/4 beam. Earth view data will be able to be corrected for this effect. There is no plan for a hardware fix to eliminate this effect.
Bob Lee of NASA/Langley Research Center (LaRC) presented the calibration status report for the Clouds and the Earth's Radiant Energy System (CERES) instruments. Lee reported no show stoppers with respect to the CERES instruments and their calibration. The CERES instrument for the Tropical Rainfall Measuring Mission (TRMM) was delivered to NASA/GSFC in October 1995 and successfully integrated in January 1996. The AM-1 CERES calibration is scheduled for completion in September 1996. Lee reported that the new CERES instruments do not exhibit the 20 count offsets of the Earth Radiation Budget Experiment (ERBE) instruments. The CERES instrument is calibrated prelaunch in a hard vacuum environment. On orbit, CERES plans to perform calibrations initially at high frequency then decreasing with time. Data validation techniques will not be sufficient to verify the CERES on-board calibration. On AM-1 the two CERES instruments plan to use series of common cross-over points to perform sensor cross-calibration. Lastly, the CERES concern that the deployment of diffuse calibration panels by the Multi-angle Imaging Spectro-Radiometer (MISR) instrument on the AM-1 platform interferes with the field of view of CERES has been resolved.
Carol Bruegge of the Jet Propulsion Laboratory (JPL) presented the calibration status report for MISR. Bruegge reported that the MISR camera calibrations were being finalized. In that process several sets of bulbs have been installed and replaced in the MISR integrating sphere. During the course of the calibration, several interesting problems were discovered and either have been or are being addressed. These items include the following: (1) strong, strategically positioned sources causing anomalies in the MISR optical performance, including saturated pixels affecting neighboring pixels; (2) illumination of MISR channel stops causing erroneous digital number (DN) output; (3) MISR performance when viewing dark targets in bright scenes; and (4) a 3% out-of-band contribution to certain MISR in-band signals. Bruegge reported that the out-of-band contribution will be corrected to 0.5% or better using MISR channel reflectance retrievals. The implications of this correction for MISR data processing will be determined by the MISR processing team and communicated to the EOSDIS.
Bill Barnes of NASA/GSFC presented the calibration status report for the Moderate Resolution Imaging Spectroradiometer (MODIS) on behalf of Bruce Guenther. Barnes reported that MODIS has experienced scattered light problems, some of which could be characterized by performing deep space and lunar spacecraft maneuvers on orbit. Barnes presented information on the variable response of the MODIS instrument with scan angle, the changes being implemented to remedy optical cross-talk problems, ghosting, and the MODIS near-field and far-field response.
Concerning MODIS near-field and far-field response, bright clouds located over a dark ocean will adversely affect the ocean data due to scattered light contamination. The near-field response of MODIS can be measured but the far-field response is much more difficult to measure. The possibility of an on-orbit maneuver to provide MODIS a near-solar view in order to quantify the far-field response was introduced and discussed.
Jim Drummond of the University of Toronto presented the calibration status report for the Measurement of Pollutants in the Troposphere (MOPITT) instrument. Drummond stated that MOPITT is concerned with temperature gradients being introduced by perturbations to the normal Earth-viewing thermal environment. The need for a stable, thermal environment for the proper operation of MOPITT was emphasized.
Brian Markham of NASA/GSFC presented the calibration status report for the Landsat-7 ETM+ instrument. Markham provided an overview of ETM+ instrument operation and calibration. With respect to possible calibration problems, the ETM+ employs a partial aperture on-board solar calibrator similar to that which was used on the MultiSpectral Scanner (MSS). This calibrator is exposed on orbit. Markham reported that the primary focal plane has been integrated, and the cold focal plane has been integrated into the radiative cooler. System level tests will conclude in July 1996, with projected delivery of the instrument to the platform integrator at the end of calendar year 1996.
Richard Willson of Columbia University presented the calibration status report for the ACRIM instrument. ACRIM is a flight-of-opportunity instrument scheduled for launch in the 1998-1999 time frame. Willson emphasized the importance of developing a precise overlap strategy to maintain continuity in the total solar irradiance database. It was also suggested that ACRIM instruments should engage in comparison exercises with those existing absolute cryogenic radiometers currently being used as standards.
Bill Chu of NASA/LaRC presented the calibration status report for the SAGE III instrument. The SAGE III preliminary design review (PDR) was held in July 1995, and the critical design review (CDR) was scheduled for August 1996. SAGE III is scheduled for launch in August 1998 aboard a Russian Meteor 3M. The calibration issue of detector linearity was discussed. Chu stated that SAGE III detector linearity is characterized over a dynamic range which encompasses the dynamic range experienced on orbit.
George Aumann of JPL presented the calibration status report for the Atmospheric Infrared Sounder (AIRS), the Humidity Sounder from Brazil (HSB), and the Advanced Microwave Sounding Unit (AMSU). With respect to AIRS, which will fly on the PM-1 platform, Aumann expressed concern that the implementation of orbital maneuvers might upset the thermal environment of the platform and cause discontinuities in the climate record. Aumann reported that AMSU, built by Aerojet Corporation, is exceeding specifications. The HSB will probably be a copy of the National Oceanic and Atmospheric Administration (NOAA) K, L, M instrument.
Brian Johnson of the National Center for Atmospheric Research (NCAR) presented the calibration status report for the High Resolution Dynamic Limb Sounder (HIRDLS). HIRDLS desires a small pitch maneuver of the Chem spacecraft to enable HIRDLS to measure radiometric strays via a full elevation and azimuth scan of deep space. A suggestion was made during the presentation that an interferometer could possibly be used instead of a monochromator in the Oxford calibration chamber. The concern that calibration could be lost in shipment of the instrument between the United Kingdom and the United States was discussed.
Ron Holm of JPL presented the calibration status report for the Tropospheric Emission Spectrometer (TES). TES is scheduled for a 2001 launch. Holm presented an overview of the instrument and its science. Currently, the Airborne Emission Spectrometer (AES) serves as a testbed for TES. The calibration plan for TES is being developed, and the instrument will eventually be moved into the MISR cleanroom facility at JPL. The plan is to maximize the use of MISR personnel and instrumentation in the build of the TES instrument.
Gary Rottman of the University of Colorado presented the calibration status report for the SOLSTICE instrument. The SOLSTICE instrument has a strong heritage with the SOLSTICE instrument which was launched on the Upper Atmosphere Research Satellite (UARS). Rottman reported that, upon opening the instrument, the UARS SOLSTICE experienced a 10% degradation over 5 years. The importance of maintaining and operating beam line 2 at the NIST Synchrotron Users Research Facility (SURF) in support of the calibration of SOLSTICE was emphasized.
Carol Johnson of NIST presented preliminary results on the February 1995 EOS Radiometric Measurement Comparison held at NEC, Yokohama, Japan on the ASTER and Ocean Color and Temperature Sounder (OCTS) integrating spheres. Participants in the comparison included NIST, the University of Arizona Optical Sciences Center Remote Sensing Group, NASA/GSFC, and the National Research Laboratory of Metrology (NRLM). Radiometric measurements were made by NEC on the two spheres the week before the formal comparison. Comparison of the percent differences between the comparison participants' measurements and the NEC measurements indicated agreement to better than 2 to 3% for 3 light levels on the ASTER sphere and 4 light levels on the OCTS sphere.
Following her presentation on the February 1995 Radiometric Measurement Comparison, Carol Johnson presented detailed information on the EOS Visible Transfer Radiometer (VXR) and plans for an EOS Shortwave Infrared Transfer Radiometer (SWIRXR). The EOS VXR follows the optical and electronic design of the SeaWiFS Transfer Radiometer (SXR). The EOS VXR is reported to be assembled and ready for characterization at NIST. The design for the SWIRXR is about to begin.
Joe Rice of NIST presented detailed information on the EOS Thermal Infrared Transfer Radiometer (TIRXR). The TIRXR is a portable radiometer which can operate in ambient or vacuum environment and can be used to verify the radiance scale of the blackbody sources used in the thermal infrared calibration of EOS instruments. Rice indicated that the TIRXR will be absolutely calibrated against the NIST Low Background Infrared Facility's (LBIR) Absolute Cryogenic Radiometer (ACR) using a large-aperture transfer blackbody. The accuracy of that calibration is anticipated to approach 1% in absolute radiance. In addition, the TIRXR will be used to place EOS instrument calibration blackbodies on a relative radiance temperature scale to 20 mK. Rice presented a number of compatibility and interface issues which must be addressed by the EOS instrument calibration facilities and NIST in advance of deployment of the TIRXR. The possibility of sending EOS calibration blackbodies to NIST for calibration was discussed.
Stuart Biggar of the University of Arizona made a presentation on their visible/near infrared and shortwave infrared transfer radiometers. The University of Arizona visible/near infrared radiometer operates from 0.4 to 0.9 um using a silicon trap detector. The shortwave infrared radiometer operates from 0.7 to 2.5 um using a cooled indium antimonide detector. Both radiometers use a series of interference filters for wavelength channel selection. The visible/near infrared and shortwave infrared radiometers will participate in the August radiometric measurement comparison at Santa Barbara and Pasadena, CA.
Jim Butler of NASA/GSFC presented information on the NASA/GSFC Scanning Spectroradiometer. The scanning spectroradiometer transfers the irradiance scale from an irradiance standard lamp to the integrating sphere under test. A knowledge of the geometry of the measurement of the integrating sphere is then used to calculate the sphere radiance. Butler outlined a number of improvements in the NASA/GSFC equipment which, upon implementation, should improve the efficiency of the NASA/GSFC measurements in the EOS radiometric measurement comparisons.
Jim Butler outlined plans for future radiometric measurement comparisons. In August, a comparison is planned to perform radiometric measurements on the MODIS and Landsat-7 ETM+ integrating spheres at Hughes Santa Barbara Remote Sensing and on the MISR sphere at the Jet Propulsion Laboratory. In addition, the second ASTER comparison has been scheduled for October 1996 in Yokohama and Kamakura, Japan. Future comparison and comparison-related activities will include revisiting the MODIS sphere in Santa Barbara, coordinating the deployment of the EOS TIRXR to EOS instrument calibration facilities, and investigating approaches for creating radiometric links to other international remote sensing instruments.
The final presentation of the first day of the meeting was by Yvonne Barnes of NIST on the status of the EOS artifact bidirectional reflectance distribution function (BRDF) round-robin. Barnes indicated that the purpose of the round-robin was to circulate a set of common reflectance targets among EOS instrument calibration laboratories and other metrology laboratories in an effort to quantify the anticipated spread of diffuse reflectance measurements. The round-robin approach will employ a set of 4 targets to be measured by 6 laboratories at a number of wavelengths in the visible/near infrared and shortwave infrared wavelength region. The NIST Spectral Tri-Function Automated Reference Reflectometer (STARR) facility will serve as the hub institution during the round-robin.
Meeting Day 2: July 10, 1996
The first presentation on the second day of the meeting was by Phil Slater. Slater provided a preliminary report on the first vicarious calibration cross-comparison campaign held in May/June 1996 at Railroad Playa and Lunar Lake in Nevada. Slater stated that the purposes of the comparison were: (1) to compare top-of-the-atmosphere spectral radiances at specified monochromatic wavelengths and selected passbands in the solar reflective range, and (2) to provide an opportunity to conduct a variety of infrared experiments. The comparison also provided an excellent indication of the robustness of the vicarious calibration techniques. Slater stated that the preliminary results of the campaign are being analyzed now, and that reports will be completed by the participating groups by the end of July. Slater suggested that July would be a good time to begin to solicit suggestions from potential participants in a 1997 campaign, with a goal of October 1996 for finalization of plans. Slater also emphasized the important role that vicarious calibration will play in the EOS program. Slater estimated that there are 30-to-40 vicarious results being produced by more than a dozen research groups world-wide. The work of these groups should possibly be coordinated by the EOS Calibration Scientist in order to maximize global calibration science benefits.
The remainder of the morning was devoted to presentations by each AM-1 instrument on plans for validation of their Level 1 (i.e., radiance) products. Phil Slater began by presenting information on the role of vicarious calibration in the ASTER Level 1-B product. Slater stated that the U.S. and Japanese ASTER teams have extensively discussed how to use vicarious calibration results with respect to the ASTER radiance product. He said that an approach for incorporating vicarious results into the Level 1 product is to convene a meeting of a review panel of specialists in pre-flight round-robin activities and vicarious calibration, science team members, and instrument experts. The weighing of the importance of vicarious calibration results will be determined through meetings of that group. Slater also proposed the use of a newsletter to communicate changes in vicarious calibration results and their effect on the Level 1 product.
The CERES Level 1 data validation plans were presented by Bob Lee. Lee stated that CERES does not plan to change calibration coefficients based on validation results unless the coefficients exceed certain pre-flight-determined levels. Lee also stated that CERES on TRMM plans to use deep space views provided through spacecraft maneuvers to quantify their DC offset.
Carol Bruegge stated that the source of MISR vicarious calibration radiances will be the MISR validation team. The MISR approach will be to decide up-front on the criteria for updating Level 1B calibration coefficients and then automatically to implement those updates. Bruegge also reported that MISR will use the high-altitude instrument, AirMISR, to perform vicarious calibration.
Bill Barnes presented the MODIS plans for validating the Level 1B product. Barnes stated that the MODIS Science Team wants the Level 1B algorithm to remain constant. The recommendation was made from the panel that since the international community wants these data in a timely manner, MODIS should be prepared to implement any changes to the calibration coefficients in the first quarter of 1999, as dictated by Level 1 validation results. Barnes also presented an indexing scheme to reflect the quality/uncertainty of MODIS data at the pixel level.
Jim Drummond and Mark Smith presented the MOPITT plans for validating their Level 1B data product. Drummond stated that the MOPITT Level 1B product is an intermediate one, in the sense that it is a feed to the Level 2 processing. Drummond also stated that MOPITT is searching for other sources of methane concentration data for validation purposes, such as the data from the Network for the Detection of Stratospheric Change (NDSC). Smith presented information on the two MOPITT aircraft instruments: the MOPITT Algorithm Test Radiometer (MATR) and MOPITT Airborne (MOPITT-A). A desire was expressed for MATR to be included in the list of potential participants in future vicarious calibration campaigns.
Phil Slater led a discussion of plans for future vicarious calibration campaigns. Slater proposed round-robin laboratory BRDF measurements of large diffuse panels in support of the reflectance-based field measurements. A complementary or alternative approach would be to perform an in-field comparison of diffuse panels and radiometers employing the sun as a common light source. Slater pointed out that similar approaches would work for radiance-based measurements. Thermal infrared radiometers could be compared through cross-comparison measurements on accompanying blackbodies. Agreement between vicarious calibration results depends on detailed description of each participant's procedures, careful characterization of instruments and artifacts, and the establishment of a set of common, accepted measurement, data analysis, and reporting protocols. Slater indicated that the adoption of a single solar spectral exo-atmospheric irradiance scale by comparison participants would contribute to reducing measurement uncertainties. Slater stated that a subgroup of the EOS Calibration Panel should be formed under the leadership of the EOS Calibration Scientist to organize the next vicarious calibration campaign.
An overview of the SeaWiFS Intercalibration Round-Robin Experiment (SIRREX) was given by Carol Johnson. The purpose of SIRREX is to transfer the NIST scale of spectral irradiance through NASA/GSFC to all participating national and international laboratories in the SeaWiFS ocean color community and to the standards used to calibrate the SeaWiFS instrument for radiance responsivity. Johnson stated that crucial to this process was the formulation by the SeaWiFS ocean color community of a series of protocols to define validation instrument setting parameters and measurement protocols.
Ambler Thompson of NIST presented information on the North American UltraViolet-B (UV-B) Intercomparison Program. Thompson presented information on the participating instruments and their measurements obtained during the UV-B intercom-parisons held at Table Mountain, Colorado.
Jim Butler led a discussion on strategies for the extension of the EOS calibration program to EOS validation instruments. The discussion was not limited just to validation instruments involved in vicarious calibration but also included instruments used in the validation of higher order data products. Butler led discussions on a number of related topics in this area including defining the proper balance between measurement and review and assessing program budgetary and technological feasibility. Butler's proposed strategy separated the validation community into the vicarious calibration community and the higher order data product validation community. The formation of working groups in each community was proposed with the higher order data product validation working groups involving representation across all EOS instrument teams, broadly organized along the lines of the EOS 24 Measurement Sets: Atmosphere, Solar, Land, Ocean, and Cryosphere. The recommendation was made by the Calibration Panel that higher order data product validation and vicarious validation activities not be separated so rigorously.
Carol Bruegge led the final discussion of the day on radiance and reflectance products in EOS. Bruegge presented three areas in which a uniformly adopted approach by EOS will facilitate the exchange, use, and comparison of results between EOS instruments. These areas included the following: the effect of source color temperature and accounting for its effects in the form of instrument out-of-band contributions, the calculation of band-averaged radiance versus monochromatic radiance, and the adoption of a common solar irradiance spectrum.
Meeting Day 3: July 11, 1996
The first portion of this final half-day session was devoted to presentations on two calibration techniques. The first technique, presented by Phil Slater, was Solar Radiation Based Calibration (SRBC) as an example of a unified approach to the pre- and in-flight, full-aperture calibration of satellite sensors. Slater clearly identified several areas in SRBC where protocols could be established for defining and standardizing the technique. These areas included measurement of the BRDF of solar diffusers, calibration of radiometers, calculation of the solar aureole correction, agreement on exo-atmospheric spectral irradiance values, characterization of the mirror used to reflect solar irradiance into the instrument under pre-flight calibration, and use of an EOS transfer radiometer to determine the radiance of the diffuser at the time of the SRBC. Stuart Biggar followed with a detailed error budget for the SRBC technique.
Hongwoo Park of NASA/GSFC presented an overview of the Total Ozone Mapping Spectrometer (TOMS) instrument, its prelaunch calibration, and the backscattered ultraviolet (BUV) measurement approach. Park provided information on the prelaunch irradiance, radiance, and goniometric calibration of the TOMS instrument and their effective use of multiple calibration approaches.
Jim Butler and Greg Hunolt, both of NASA/GSFC, presented information on archiving calibration data. Butler, in his presentation, identified the following eight sources of archivable calibration data: radiometric measurement comparison data, artifact round-robin data; Level 1 validation data, pre-flight instrument calibration data, on-orbit instrument calibration data; data from the calibration of higher order data product validation instruments, image-based analysis data, and cross-comparison satellite data. Hunolt stated that archiving calibration data is recognized by EOSDIS as a fundamental requirement for science support. Calibration data will be archived at the Distributed Active Archive Centers (DAACs), and these data will be accessible from EOSDIS Core System (ECS) data servers. Hunolt charged the instrument teams to define those calibration data which need to be archived to ensure that scientists have all the information needed to understand the data. The instrument teams then need to coordinate this information with the responsible DAAC. Hunolt suggested that the Science Software I&T Procedures could possibly be where the specific archiving requirements for these calibration data should be documented.
James Anderson of Northern Arizona University (NAU) presented an overview of the lunar radiometric measurement program being conducted jointly by NAU and the United States Geological Survey (USGS) in Flagstaff, Arizona. This is a multi-year program in which Earth-based observations of the Moon are accumulated, corrected for atmospheric extinction using a series of standard star observations, and calibrated using an in-dome integrating sphere and standard lamp observations. The calibrated and corrected lunar images are used to produce an exo-atmospheric lunar radiometric model for each lunar pixel. The lunar radiometric model will be used by those EOS instruments which are able to view the Moon in the on-orbit determination of those instruments' visible, near infrared, and shortwave infrared radiometric responsivities.
Jim Butler concluded the calibration meeting with a discussion of EOS platform maneuvers for deep space and lunar viewing. The five-step MODIS roll-based maneuver and the MISR/ASTER five-step pitch-based maneuver were illustrated and discussed. The five-step maneuvers satisfy the deep space and/or lunar viewing desires of MODIS, CERES, MISR, and ASTER. However, in order to minimize thermal impact on the MODIS passive cooler and to provide ASTER with a view of the Moon between 4 and 11 degrees lunar phase, the two maneuvers must be performed on orbits separated by several days. This is contrary to the desire of the EOS AM-1 project to present these maneuvers as a single maneuver performed on consecutive orbits. Butler warned that a discussion of lunar/deep space viewing requirements needs to begin for the PM-1 platform with the formation of a calibration attitude maneuver (CAM) working group for the PM-1 platform.