Calibration in the EOS Project
Part 2: Implementation
--James J. Butler (butler@highwire.gsfc.nasa.gov), EOS Calibration Scientist, NASA/Goddard Space Flight Center
--B. Carol Johnson (cjohnson@enh.nist.gov), Optical Technology Division, National Institute of Standards and Technology
To fulfill its mission, EOS must produce accurate, precise, and consistent long-time series measurement data from multiple instruments on multiple platforms. The correct interpretation of these long-time series data requires the ability to differentiate actual changes in the remotely sensed Earth from changes in the measuring instrumentation. This can only be accomplished by (1) calibrating all instruments against a set of recognized physical standards, (2) carefully characterizing the instruments' performances at the system level, (3) adhering to good measurement practices and established protocols, (4) intercomparing instrument measurements, and (5) establishing traceability to the physical standards via an impartial standards organization.
This article constitutes Part 2 of a two-part article describing the overall organization and implementation of a calibration program in the EOS Project based on requirements initially established in 1989 (EOS Project Requirements Level 1A, 1989). Part 1 (Butler and Johnson 1996) described the organization of the EOS Calibration Program, its position in the EOS Project Science Office's Panel for Data Quality, and the implementation of the program with respect to planning, documentation and peer reviews. This article completes the description of the implementation of the program by outlining the EOS approaches to pre-flight and on-orbit calibration. Key to these approaches are the measurement assurance programs (MAPs) in pre-flight and on-orbit calibration supported by the National Institute of Standards and Technology (NIST), the lunar radiometric measurement program for on-orbit calibration being conducted by the United States Geological Survey (USGS) and Northern Arizona University (NAU), and the execution of field programs to validate EOS instrumental Level 1B data. Where appropriate, examples of on-going calibration programs relevant to the EOS AM-1 instruments are provided.
Figure 1 illustrates the implementation of the EOS Calibration Program and shows the positions of instrument calibration and cross-calibration in the overall implementation scheme (the term cross-calibration is defined below). Pre-flight calibration and pre-flight and on-orbit cross-calibration of EOS instruments are critical to the success of the EOS mission. Extension of these calibration and cross-calibration activities to field instruments involved in the Level 1B data validation of the EOS instruments ultimately improves the validity and reliability of the EOS instruments Level 1B data. As seen in Figure 1, EOS calibration and cross-calibration is a multifaceted program, incorporating the NIST MAPs, the USGS/NAU lunar radiometric measurement program, and Level 1B data validation field programs.
A MAP is any group of activities designed to critically evaluate the accuracy of a group of measurements. When implemented by NIST, it is a quality control procedure designed to calibrate a customer`s entire measurement system (Simmons 1991) and establish traceability to NIST. A key part of the EOS Calibration Program are the MAPs operated by the EOS Project Science Office with support from NIST. In addition to evaluation of the pre-flight and on-orbit instrument calibration activities, the MAPs include:
As indicated in Figure 1, the components of the MAPs include the following: radiometric measurement programs, artifact round-robins, other measurement comparisons/services, and intercomparisons, workshops, and training. These are examined below.
In the pre-flight timeframe, EOS instrument calibration facilities use dedicated, large-area, calibrated sources, i.e., integrating spheres, blackbodies, etc., to determine the radiometric response of instruments operating at optical wavelengths, i.e., ultraviolet to thermal infrared. With the EOS instruments acting as transfer radiometers, these well-characterized, large-area sources are usually used to establish the radiometric scales of any on-board radiometric sources. It is extremely important that the calibration of the laboratory standard sources be consistent between the EOS instrument calibration facilities and accurate with respect to System International (SI) units.
The objective of the radiometric measurement programs is the pre-flight verification of the independent radiometric scales assigned to the sources used in the actual calibration of the EOS instruments. These measurements are usually made at the EOS instrument calibration facilities. The approach is to use stable, portable, well-characterized radiometers traceable to a national standards laboratory to measure the EOS calibration sources. NIST is in the process of designing, building, characterizing, and deploying a set of radiometers that will be used to verify the spectral radiance of these EOS instrument calibration sources. Three radiometers are planned: the EOS Visible Transfer Radiometer (VXR), a six-channel filter radiometer based on the Sea-Viewing Wide Field-of-view Sensor (SeaWiFS) Transfer Radiometer (SXR) (Johnson et al. 1996a), the EOS Shortwave Infrared Transfer Radiometer (SWIXR), a modification of the EOS VXR for the shortwave infrared region, and the EOS TXR, a two-channel cryogenic filter radiometer that can be operated in vacuum or ambient conditions (Rice and Johnson 1996). These radiometers may be deployed with small, stable sources in order to monitor their performance. In addition to verifying the spectral radiance of critical radiometric sources, it is anticipated that these radiometers will be used to perform a number of other functions, including use in validation field programs and measurement technique comparisons. These specific applications are discussed below in the sections entitled Validation Field Programs and NIST Intercomparisons, Workshops, and Training. In fiscal year 1996, NIST will deliver to the EOS Project Science Office the VXR, the engineering plan for the SWIXR, and test data on the TXR prototype.
The radiometric measurement programs have begun. The initial deployment was held at NEC in Yokohama, Japan, in February 1995. During that activity, the SXR, operating in the visible and near-infrared wavelength regions, viewed the integrating spheres used to calibrate the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and the Ocean Color and Temperature Scanner (OCTS). In addition to the SXR, radiometers from the National Research Laboratory of Japan (NRLM) (Sakuma et al. 1994), the University of Arizona Optical Sciences Center (Biggar and Slater 1993), and a scanning spectroradiometer from NASA Goddard Space Flight Center (Johnson et al. 1996a) also viewed the spheres. Preliminary results indicate an agreement between these measurements of better than 3% for three radiance levels of the ASTER sphere (compared to program requirements of 4%) and four radiance levels of the OCTS sphere (Sakuma et al. 1996, Johnson et al. 1996b). The next measurements are being planned for August 1996, during which time the radiometric scales for the integrating spheres used to calibrate the Moderate Resolution Imaging Spectro-radiometer (MODIS), Enhanced Thematic Mapper Plus (ETM+), and the Multi-angle Imaging Spectroradiometer (MISR) will be verified.
These radiometric measurements at instrument calibration facilities using the EOS/NIST radiometers will be performed throughout the entire EOS mission lifetime. Only by performing comprehensive and thorough comparison measurements is it possible for the EOS program to generate long-term, continuous, consistent data that can be confidently subjected to critical evaluation. The practical difficulties of designing and deploying a common, stable, radiometric source deployed at the platform integration facility and viewed by the integrated EOS instruments led to the transfer radiometer approach described above (Minutes of the Sixth General Meeting of the EOS IWG Calibration Panel 1993). Instrument teams are encouraged to bring sources and equipment to the platform integration facility for purposes of performing bench acceptance testing and ensuring that instrument calibration has not changed. The nature of the sources, accompanying equipment, and planned tests are communicated by each instrument team to the EOS Project and the platform integrator well in advance of instrument shipment and platform integration.
A second activity performed by NIST in support of the EOS Calibration Program is to circulate samples for measurement at selected laboratories. For example, visible, near infrared, and shortwave infrared EOS instruments require an accurate determination of the Bi-directional Reflectance Distribution Function (BRDF) of on-board and laboratory diffuse plaques. Many remote sensing instruments and calibration facilities use diffuse plaques to:
In order to address the scientific goals of the EOS program, measurements of BRDF at state-of-the-art accuracies are required. The simplest way to assess the quality of BRDF measurements is to ask a number of EOS instrument calibration and metrology laboratories to measure the same samples. This EOS-sponsored BRDF round-robin is underway, with the Spectral Tri-function Automated Reference Reflectometer (STARR) facility at NIST serving as the hub in the measurement program (Proctor and Barnes 1996). Other participating laboratories include the University of Arizona Optical Sciences Center, Rochester Institute of Technology, GSFC, Hughes Santa Barbara Remote Sensing (Hughes SBRS), and the Jet Propulsion Laboratory (JPL). Four samples will be measured, consisting of laboratory grade Spectralon(TM) from Labsphere, Inc.+, pressed polytetrafluoroethylene (PTFE), baked PTFE, and diffuse aluminum from Ball Aerospace. The PTFE samples will be provided by NIST.
Additional candidate round-robin programs may be identified by members of the EOS Calibration community, presented to the community at EOS Calibration Panel meetings, and brought to the attention of the EOS Calibration Scientist in the form of a well-conceived test plan. For example, round-robin programs for dimensional metrology, e.g., aperture area, and spectrophotometry, e.g., filter transmittance are also possible. NIST is available as the hub, since facilities exist for making accurate measurements of the optical area of apertures (Fowler and Dezi 1995) and regular spectral transmittance measurements on room temperature (Mielenz 1973) and cryogenic filters (Datla, pers. comm.) As with the BRDF round robin, these activities will serve to corroborate the measurement practices, instrumentation, and capability of EOS-related efforts at the various laboratories and commercial facilities; in general they will not be simultaneous with calibration and characterization of flight hardware.
For the past four years, NIST has participated in intercomparisons in support of terrestrial ultraviolet (UV) spectral irradiance and ocean color science. The valuable experience gained in these activities will be applied to EOS. The 1994 UV Interagency Intercom-parison, led by NIST and the National Oceanic and Atmospheric Administration (NOAA), emphasized instrument characterization, in situ interinstrument calibration, and sun-synchronous measurements (Thompson et al. 1996). The program was repeated in 1995 and is scheduled for a third implementation in June of 1996. The ocean color work has been led by the Calibration and Validation Program of the SeaWiFS Project Science Office at GSFC (McClain et al. 1992). Beginning in 1992, based on protocols developed in 1990 by the science community (Mueller and Austin 1992), a series of SeaWiFS Intercalibration Round Robin Experiments (SIRREXs) have been held annually, with NIST participation. The Fourth SIRREX emphasized training and workshops (Johnson et al. 1996a), while the Fifth, scheduled for July 1996, will concentrate on in-water spectral radiance and irradiance measurements, field reflectance measurements, and methods to realize spectral radiance scales for calibration of field radiometers. A general result of the UV intercomparisons and the SIRREXs is that radiometric instruments often suffer from insufficient characterization, inadequate design or improper use when evaluated with respect to the uncertainties required by the underlying science program. Attention to these areas resulted in improved results, a better-educated community, and specific radiometric artifacts for transfer of NIST scales to the science community (Johnson et al. 1996a).
These activities will be extended to the EOS community. Over the life of the EOS mission, the performance of the various sources and radiometers used not only in the calibration of EOS and other Mission to Planet Earth (MTPE) instruments but also those used in field programs must be evaluated. The NIST/EOS transfer radiometers and other NIST/EOS radiometric equipment, e.g., sources, will be used to assess the accuracy of the radiometric calibration of the field equipment, and place all instruments in the intercomparison on a common scale so that performance issues can be addressed independent of radiometric calibration. It will then be possible to ascertain if instruments of different design, manufacture, and calibration method give the same result under similar measurement conditions. As appropriate, workshop and training activities will be included in the intercomparisons. The entire effort will be organized along the classic wavelength disciplines of visible/near infrared, shortwave infrared, thermal infrared, and microwave. The workshop activities will include presentations on instrument calibration and characterization, demonstrations of good radiometric/calibration technique, comparisons with actual field radiometers, and reviews of the results of the EOS Calibration/Validation efforts.
At the USGS and at NAU, both in Flagstaff, Arizona, long-term radiance measurements of the moon are being made in support of the on-orbit calibration, cross-calibration, and characterization of EOS visible and shortwave infrared EOS instruments (Kieffer and Wildey 1996). The moon is the only object accessible to all terrestrial orbiting spacecraft that is within the dynamic range of most imaging instruments and is stable enough to provide a calibration target. In Flagstaff, the USGS and NAU have constructed a dedicated observatory housing a visible/near infrared imaging telescope/camera system. Later this year, a shortwave infrared imaging system will be added to the observatory, extending the spectral radiometric measurement capability to 2.5 um. These systems will make observations of the moon and sets of standard stars every photometric night over the bright half of the month over at least a 4.5-year continuous interval. USGS and NAU will use these data to construct a lunar radiometric model that will generate accurate exoatmospheric radiance data corresponding to EOS spacecraft instrument observations. The EOS/NIST and University of Arizona radiometers described earlier in this article will be used to verify the spectral radiance scale of the lunar radiance measurement calibration equipment.
When the field-based and space-based measurements in the Level 1B data validation program are spatially and temporally simultaneous and the sensors are similar, the field measurements can be used to validate or verify the radiometric calibration coefficients of the space-based sensor. The EOS Calibration and Validation Scientists recognize the importance of promoting well-organized Level 1B data validation field programs in support of the EOS mission. In an effort to promote and foster these programs, the EOS Calibration and Validation Scientists and the EOS Deputy Senior Project Scientist are continuing the effort started by the Committee on Earth Observation Satellites Working Group on Calibration and Validation (CEOS/WGCV) (CEOS Pilot Cal/Val Dossier 1993). The goal of this effort is to produce a reasonably detailed international database of calibration facilities, test sites, and field instruments. This information will ultimately be used to identify participants for EOS calibration programs and common field test sites.
In order to improve the overall quality of the EOS instrument Level 1B data, validation field programs will begin in advance of spacecraft launch and will ultimately produce from each participant a representative, long-time series data set. In advance of the field programs, participants will: (1) clearly establish measurement, data analysis, and data reporting protocols, (2) fully characterize their instruments, and (3) calibrate their instruments using methods and artifacts traceable to a national standards laboratory.
In order to maximize the benefit to the EOS instruments, the calibration of Level 1B data validation instruments should take place before and after the field programs. In the case of radiometers, these pre- and post-calibrations should involve the near-simultaneous viewing of well-characterized radiance sources in a controlled laboratory environment.
The implementation of calibration and cross-calibration in the EOS Project includes a number of measurement programs. The NIST-supported MAPs in pre-flight and on-orbit calibration and cross-calibration are a key component of calibration in EOS. The MAPs currently include radiometric measurement programs, artifact round robins, measurement comparisons, workshops, and training. A measurement program in support of the on-orbit calibration and cross-calibration of EOS/MTPE and international sensors is the lunar radiometric characterization being performed by USGS and NAU in Flagstaff, Arizona. The Level 1B data validation programs are a third measurement activity which will verify the on-board calibration systems of orbiting EOS instruments.
Biggar, S. F., and P. N. Slater, 1993: Preflight cross-calibration radiometer for EOS AM-1 platform visible and near-IR sources, SPIE Proc., 1939, 243.
Butler, J. J., and B. C. Johnson, 1996: Organization and implementation of Calibration in the EOS Project -- Part 1, The Earth Observer, 8, 22-27.
Committee on Earth Observation Satellites (CEOS) Pilot Cal/Val Dossier, Smith System Engineering, Ltd., Guildford, UK, October 1993. Earth Observing System (EOS) Project Requirements Level 1A, pp. 15-16, May 10, 1989.
Fowler, J., and G. Dezi, 1995: High accuracy measurement of aperture area relative to a standard known aperture, J. Res. NIST, 100, 277-283.
Johnson, B. C., S. S. Bruce, E. A. Early, J. M. Houston, T. R. O'Brian, A. Thompson, S. B. Hooker, and J. L. Mueller, 1996a: The Fourth SeaWiFS Intercalibration Round-Robin Experiment, SIRREX-4, May 1995. NASA Tech. Memo. 104566, 35, S. B. Hooker and E. R. Firestone Eds., NASA Goddard Space Flight Center, Greenbelt, Maryland, 78 pp.
Johnson, B. C., F. Sakuma, S. F. Biggar, J. Butler, J. Cooper, M. Hiramatsu, and K. Suzuki, 1996b: OCTS-SeaWiFS preflight cross-calibration experiment at the OCTS integrating sphere using round robin radiometers. Publication in preparation.
Kieffer, H. H. and R. L. Wildey, 1996: Establishing the moon as a spectral radiance standard, J. Atmos. Oceanic Tech. 13, 1996, 360-375 pp.
McClain, C. R., W. E. Esaias, W. Barnes, B. Guenther, D. Endres, S. Hooker, G. Mitchell, and R. Barnes, 1992: Calibration and Validation Plan for SeaWiFS, NASA Tech. Memo. 104566, 3, S. B. Hooker and E. R. Firestone, Eds., NASA Goddard Space Flight Center, Greenbelt, Maryland, 41 pp.
Mielenz, K. D., 1973: Physical Parameters in High-Accuracy Spectrophotometry, NBS Spec. Pub. 378, National Bureau of Standards, Washington, DC, 81-93.
Mueller, J. L. and R. W. Austin, 1992: Ocean Optics Protocols for SeaWiFS Validation, NASA Tech. Memo. 104566, 5, S. B. Hooker and E. R. Firestone, Eds., NASA Goddard Space Flight Center, Greenbelt, Maryland, 43 pp.
Proctor, J., and Y. P. Barnes, 1996: NIST High Accuracy Reference Reflectometer-Spectrophotometer. accepted for publication in J. Res. NIST., 101.
Rice, J. P., and B. C. Johnson, 1996: Design of a Thermal Infrared Radiometer as a Transfer Standard in the NASA EOS Program. In preparation for submission to J. Res. NIST.
Sakuma, F., M. Kobayasi, and A. Ono, 1994: ASTER round-robin radiometers for the preflight cross-calibration of EOS AM-1 instruments, Proc. IGARSS, 1995-1997.
Sakuma, F., B. C. Johnson, S. F. Biggar, J. Butler, J. Cooper, M. Hiramatsu, and K. Suzuki, 1996: EOS AM-1 pre-flight cross-calibration experiment at the ASTER VNIR integrating sphere using round-robin radiometers. Publication in progress.
Simmons, J. D., Ed., 1991: NIST Calibration Services Users Guide 1991, NIST Special Publication 250, U.S. Department of Commerce, Gaithersburg, Maryland, 200 pp.
Thompson, A., E. A. Early, J. DeLuisi, P. Disterhoft, D. Wardle, J. Kerr, J. Rives, Y. Sun, T. Lucas, M. Duhig, and P. Neale, 1996: The 1994 North American interagency intercomparison of ultraviolet monitoring spectro-radiometers. Submitted to J. Res. NIST.
| Table of Contents | | Previous | | Next |
