The Landsat 7 Science Team held their semi-annual team meeting December 1-3 at the NASA Goddard Space Flight Center (GSFC). Participants from the 14 research teams, as well as NASA, NOAA, and USGS, attended the meeting, which was co-chaired by Landsat 7 Science Team Leader, Samuel Goward, and Landsat Project Scientist, Darrel Williams.
Following welcoming and farewell remarks from Tony Janetos (NASA HQ Landsat Program Scientist), briefings were given on the status of the Landsat platform and ETM+ instrument by the Landsat Project Manager, Phil Sabelhaus (GSFC), and on the Landsat Ground System by Deputy Project Scientist, Jim Irons (GSFC). Sabelhaus announced that a launch date of April 15, 1999 had been set, but there was a chance that the date might be pushed ahead to March 31, 1999. Although the original delivery of the ETM+ instrument was delayed due to faulty power supplies, the problem has now been corrected, and work was on schedule for either the late-March or mid-April launch date. Sabelhaus did note some minor problems that emerged during radiometric and geometric testing, and indicated that these issues would be addressed during the next few months. Irons reported that the Landsat 7 Ground System is nearing completion and is on schedule for the launch and the subsequent transition to nominal operations. The Landsat Mission Operations Center (MOC) at Goddard Space Flight Center is launch ready. The Flight Operations Team (FOT) and the Flight Support Team (FST) are preparing for launch by participating in a series of "launch and on-orbit activation" simulations in the MOC. The Landsat Data Handling Facility (DHF) is in place at the U.S. Geological Survey’s (USGS) EROS Data Center (EDC) at Sioux Falls, SD. The components of the DHF (the Landsat Ground Station, Landsat Processing System, and the Image Assessment System) are undergoing final tests in preparation for launch. The ETM+ data received and processed to Level 0R by the DHF will be sent to the EDC Distributed Active Archive Center (DAAC) for archiving, processing to Level 1 data products, and distribution to ETM+ data users. Delays in the development of the EOSDIS Science and Data Processing Segment remain a concern, but recent EDC DAAC tests indicate that the Segment will be sufficiently developed to support the on-orbit check-out of the ETM+ during the first 60 days following launch.
Shortly before the Science Team meeting, representatives of NASA, NOAA, and the USGS agreed to transfer responsibility for Landsat 7 operations from NOAA to USGS. Thus, the Landsat program will effectively become a two-agency partnership between NASA and USGS beginning in 1999. James Ellickson (NOAA) and R.J. Thompson (USGS/EDC) discussed this transition. Following the 60-day spacecraft checkout period, USGS will be responsible for ground data systems support, data acquisition, and interactions with International Ground Stations (IGS). NASA will continue to be responsible for flight operations until October 2000, at which time that responsibility will be transitioned to the USGS as well. Thompson stressed that USGS will build on satellite operations expertise of both NASA and NOAA, and that USGS was committed to making Landsat 7 a success. Thompson also mentioned that the current pricing structure for Landsat 7 data may be reviewed, to explore a more favorable pricing policy for either data consumers or the IGS network.
Brian Markham, John Barker, and Jeff Pedelty (GSFC) presented a brief summary of ETM+ radiometric and geometric test data. Although the ETM+ instrument performance has generally met specifications, some anomalies did emerge during testing. These include:
These results were discussed more thoroughly during an evening breakout session presented by Markham, Barker, and Pedelty, and attended by nearly all team members. It should be stressed that none of these anomalies in instrument performance is serious enough to warrant a delay in launch, and, in nearly all respects, the radiometric and geometric performance of ETM+ appears to be superior to any other Thematic Mapper instrument from the Landsat series.
The second day commenced with a discussion of follow-on options for Landsat 7. Stephen Ungar (GSFC) presented an update on the New Millennium Program (NMP) Earth Observing 1 (EO-1), a technology demonstration targeted for launch in December 1999. The mission seeks to exploit new technology to acquire improved high-resolution observations at reduced cost. EO-1 will include a pushbroom multispectral/panchromatic array (the Advanced Land Imager — ALI), a hyperspectral imager built by TRW (Hyperion), and an atmospheric corrector sensitive to water vapor and clouds. Ungar presented test data for the ALI indicating a substantial improvement in the signal-to-noise ratio compared to the Landsat 7 ETM+ due to the greater dwell time associated with a pushbroom instrument approach.
This presentation led to a discussion of the "Landsat-Next" proposal, submitted by Darrel Williams and Jim Irons from the Landsat Project Science Office (LPSO), and Samuel Goward and Jeff Masek from the Landsat Science Team in response to last summer’s EOS Request for Information (RFI) call. Recall that in early 1998, NASA HQ solicited input via the EOS RFI to select a set of measurements for the next generation of EOS sensors. To ensure that Landsat-type measurements would continue to be a priority, the "Landsat-Next" proposal articulated the science rationale for high-resolution, seasonally acquired, multispectral observations, and outlined future measurement goals. These measurement goals include improved ground navigation of IFOVs, 12-bit radiometry, and extra spectral bands for atmospheric correction. Goward stressed that the objective of the proposal was to offer the science rationale and requirements for future high-resolution missions, but that the technical design of such missions was still open.
The afternoon of the second day revolved around a proposal from the University of Maryland to build a Landsat Science Team computing facility (REALM - Research Environment for Advanced Landsat Monitoring). Such a facility would enable science users to submit analysis tasks to a database of preprocessed Landsat data. Thus, rather than dealing with individual scenes, users could apply algorithms to arbitrary regions around the globe, allowing automated, continental-scale analyses. Goward (UMCP) noted that it had taken the Landsat Pathfinder project nearly five years to process 3000 scenes into maps of tropical deforestation. In contrast, Landsat 7 would return the same volume of data every 12 days. Such volumes of data require a drastic improvement in the way researchers handle Landsat analysis. Following Goward’s introduction, technical presentations were delivered on atmospheric correction (Guyong Wen, GSFC), automated image co-registration (Jacqueline LeMoigne, GSFC; Jim Storey, EDC), and cloud detection (Rich Irish, SSAI/GSFC). The science team agreed that such a capability would substantially benefit the Landsat science mission, and agreed to investigate options for implementation.
The day concluded with a presentation by David Herring (SSAI/GSFC), representing the EOS Project Science Office, on the new EOS Earth Observatory web site. This web site, currently under development by the EOS Project Science Office, gives the public an in-depth look at Earth System Science, and the role of satellite remote sensing. The site includes a number of innovative features, including on-line geophysical models and interactive data visualizations. Herring noted that Landsat 7 would play a major role in EOS public outreach, since it would be the first major land-imaging mission in the EOS series.
The final day of the team meeting began with a summary of the science team review of the Landsat Long-Term Acquisition Plan (LTAP) by Samuel Goward (UMCP) and Terry Arvidson (Lockheed-Martin). The LTAP provides a set of algorithms for scheduling Landsat 7 image acquisitions, with the overall goal of building a substantially cloud-free archive of global imagery documenting the seasonal cycles of Earth’s vegetation. On August 31, 1998, a one-day review of the LTAP was held at GSFC to give the Landsat Science Team a detailed view of the algorithms and their implementation. The review was successful, and all participants agreed that the LTAP represented a major contribution to the success of the Landsat 7 mission. Several minor issues were highlighted by science team members, and most of these were addressed during the December team meeting. A written summary of the LTAP is being prepared for submission to EOS Transactions, the newsletter of the American Geophysical Union.
Terry Arvidson discussed post-launch operations of Landsat 7. The first 60 days after launch will be dedicated to spacecraft health and safety, orbit stabilization, and instrument calibration. Following 18 days of outgassing, the ETM+ instrument will start acquiring test data. Concurrently, a series of burns will adjust the spacecraft orbit to meet the Landsat Worldwide Reference System (WRS) grid. This orbital drift will allow a 30-orbit underfly of Landsat-5, during which time imagery will be acquired from both platforms to aid instrument characterization. During the checkout period, the Landsat Project Science Office (LPSO) will be the sole customer for ETM+ data. Normal operations and data distribution to other users will begin after the checkout period. Richard Irish (SSAI/NASA GSFC) also presented a short overview of the data formats selected for Landsat 7 data distribution. Formats for the Level 1-G product include HDF, GeoTiff, and the EOSAT FAST format.
The meeting concluded with two presentations highlighting the use of ETM+ data for land-cover applications. Science team member James Vogelmann (Raytheon ITSS/EROS Data Center) presented an overview of the proposed More Refined Land Cover (MRLC) 2000 project. This project, a joint effort of USGS, NOAA, and the EPA, seeks to produce accurate, consistent land-cover products for the United States from Landsat TM and ETM+ imagery. MRLC 2000 would extend the current work by adding more refined land-cover classes and land-cover-change products. Jeff Morisette then gave a report on EOS Calibration/Validation activities, many of which require Landsat 7 imagery.
The meeting also included presentations from all team members on their research activities during the last six months. Readers interested in specific science team investigations should consult the article on the October 1997 Landsat Team meeting in the February/March 1998 issue of The Earth Observer. The next team meeting will be held in Solvang, California, coordinated to occur with the launch at nearby Vandenberg Air Force Base (tentative dates are April 13-14, 1999). Interested parties should contact Jeffrey Masek (jmasek@geog.umd.edu) for further information.