A paper presented at the Pecora 13 Symposium,
Sioux Falls, South Dakota,
August 22, 1996
George J. Komar
Landsat Program Manager
Office of Mission to Planet Earth
Code 170, Goddard Space Flight Center
Greenbelt, Maryland 20771
National Enviornmental Satellite, Data and Information Service
National Oceanic and Atmospheric Administration
FB-4 Room 3301E
Washington DC 20233
Senior Research Scientist
Johnson Controls World Services
NASA Ames Research Center
Moffett Field, California 94035
US Geological Survey
EROS Data Center
Mundt Federal Building
Sioux Falls, South Dakota 57198
The Landsat Program has collected images of the earth from space since 1972. Those images, as hard copy and digital data, have been used worldwide for assessment and monitoring of natural resources, land cover and land use mapping, monitoirng natural and anthropgenic disasters, global change research and education. The program has undergone significant changes in goals and management during its quarter century of operation. The next in the series, Landsat 7, will be built and launched by the National Aeronautics and Space Administration (NASA) and operated by the National Oceanic and Atmopsheric Administration (NOAA) and the US Geological Survey (USGS).
Landsat 7 will be significantly better than its predecessors. It will provide data "continuity" while offering improved calibration, spectral response and data acquisition rate. Landsat 7 is well on its way toward completion. Launch is scheduled for 1998. As required by law, the management of the Landsat Program is evaluating options for acquisition of Landsat type data in the post-Landsat 7 era. That process will consider the requirements for Landsat-type data and the changing roles of government and the commercial sector in the acquisition and distribution of earth observation data obtained from space.
While discussion continues on the long term policy for the Landsat Program, NASA has acted to maintain a robust set of options. The program has been integrated into the Earth Observing System, and space has been reserved for a Landsat-type instrument on the EOS-AM2 platform. The possibility of meeting public data requirements through data purchases is being explored, and a new program has been initiated to develop the next generation of instruments for multi-spectral imaging of the land surfaces of the earth.
Last month marked the 24th anniversary of the current era in earth remote sensing. On July 23, 1972, the first US civilian satellite was launched dedicated to acquisition of data for monitoring and assessment of earth resources. That mission proved to be successful in ways not anticipated at the time of launch. Few now remember that the primary instrument on that satellite, ERTS-1, was a device called the Return Beam Vidicom (RBV) an instrument that proved to be not nearly as capable or durable as the secondary, experimental instrument, the Multispectral Scanner System (MSS). The type of data acquired by MSS, digital, multispectral, moderate ground resolution, rectifiable, and repetitive, became the cornerstone of earth remote sensing from space. It proved its utility across a broad range of applications, from science to city planning, it became the basis for a satellite program called Landsat, and it introduced and demonstrated a technology that was, subsequently, duplicated by others.
The history of the Landsat Program has been tumultuous, reflecting uncertainty in the political, commercial and scientific sectors about how to develop and manage a new technology. But the program has been sustained. Data from Landsat 5 continues to be acquired by a global network of receiving stations. Landsat 7 is being built and will be launched two years from now. The Land Remote Sensing Policy Act of 1992 (PL102-555) established mandates and guidelines for the Landsat Program into the next century.
If the first satellite in the Landsat series marked the start of a new era, the next in the series will be launched amidst what may be the start of another. Until now, satellites collecting data about the surface of the earth have been the purview of the public sector. In some instances, government cooperated with commercial entities to build, launch and operate systems, but those programs, e.g., Landsat, SPOT, IRS, would not have started or continued without substantial government financial assistance. Consequently, those programs have been responsive to requirements established by government. Now, the commercial sector believes the market for satellite based, earth observation data has matured and the cost of building, launching and operating systems has declined sufficiently that investment in such systems will produce a satisfactory return. Although, as yet, no fully commercial system has been launched and operated to demonstrate the validity of that premise, a new era of commercial earth remote sensing from space may be upon us. The pull from the commercial sector is being supplemented by a push from government. It is clear that government, especially in the US, is looking for new ways to meet its data acquisition requirements - new ways of "doing business." The primary motivation for this interest is to reduce cost. Given the pull from the commercial sector and the push from government, the acquisition and distribution of earth observation data from space will likely be very different at the turn of the century from what it is today. As Landsat has been, and continues to be, a key element in earth observations - indeed, Landsat data (or, more accurately, Thematic Mapper data) is the data type of choice in its class within the US and internationally, it is an appropriate time to review the status of the Landsat Program and look at the options under consideration for acquisition of Landsat-type data after Landsat 7
2. Landsat 7
Details of the features of Landsat 7 were described for this symposium in a paperr by the Landsat 7 Project manager (Obenschain, 1996). Landsat 7 will offer significant improvements in data quality and data availability over that of its predecessors. In addition, the proposed flight configuration of Landsat 7 with EOS-AM1 offers potential improvements in the capabilities of both systems through complementarity .
2.1 Enhanced Thematic Mapper Plus
The instrument on Landsat 7, the Enhanced Thematic Mapper Plus (ETM+) is responsive to improvements long requested by the data user community while maintaining the essential characterisitcs of thematic mapper type data. (Table 1 (not shown)) The spectral bands present on the thematic mappers of Landsat 4 and 5 are part of ETM+. Ground resolution remains unchanged, 30 meters, except for the thermal band in which the resolution is increased from 120 meters to 60 meters. A panchromatic band with 15 meter resolution has been added for rectification and image sharpening. Of particular interest to the science community, is the addition of the capability for 5% absolute radiometric calibration - the first time for any instrument of this type. Landsat 7 will be equipped with a solid state recorder with a capacity for 100 scenes. The downlink data rate has been increased to 150 mbps and data will be transmitted to the ground via three directional antennas.
2.2 Data Acquisition
Improvements in the instrument are matched by improvements in data acqusition and availability. The US will capture 250 scenes per day at the EROS Data Center (EDC) in Sioux Falls (the primary Landsat 7 system receiving station) and a high latitude, back-up receiving station in Alaska. Included in the total is the acquisition of every sun-lit scene of the US (including Alaska and Hawaii) on every pass of the satellite. Not included in the total are the scenes collected by the network of international ground stations. Those stations will receive data by direct downlink only, i.e., they will not have access to data on the solid state recorder.
Statistically, 250 scenes is more than the number of generally cloud free scenes one can expect the ETM+ to see in a day. Consequently, Landsat 7 is capable of capturing and transmitting to the US station(s) all potential "good" scenes the instrument will see every day.
For the first time in the history of the Landsat Program, systematic collection of a sun-lit, generally cloud free, global land surface data set is part of the mission statement. This global data set will be acquired and archived at EDC. Scene acqusition requests supplementary to the mission requirements will be accepted and filled according to an established, and published, data acqusition policy. The international ground stations that acquire Landsat 7 data will likely receive every scene captured on each daylight pass of the satellite through the station's circle of coverage. As those stations will have in their archives substantially more scenes for the area within their coverage circles than will be captured by the US station, the intertnational stations will be the primary sources for coverage outside the United States.
The amount of data collected by the Landsat 7 receiving station will be approximately 4-5 times greater than the amount of data received at the existing Landsat 5 receiving station in the US.
2.3 Data Processing and Distribution
Archiving, processing, and distributing ETM+ data will also be improved significantly. All data received at the primary receiving station will be archived as level 0R. Metadata and browse files will be available for electornic user review within 24 hours of acquisition of the data. Requests for data from the archive will be filled within 24 hours. Scenes will be distributed as either level 0R or level 1G products (systematic correction applied). All users buying data through EDC will pay the cost of fulfilling user requests (COFUR). Cost for data from the IGS' will vary by station.
Several of the international ground stations have in place, or are building, on-line archive browsers for the Internet. Steps are being taken to link those browsers to the EOS on-line directory at EDC to give users, from a single point of entry to the network, ready access to the catalog information from all stations with Landsat data.
2.4 Landsat 7 Program and EOS
One of the outcomes of the re-structuring of the Landsat Program in 1994 was the integration of Landsat with NASA's Earth Observing System (EOS) program. The integration of the two programs is proving beneficial to both. At the urging of the Landsat Project scientist, Dr. Darrel Williams, a memorandum of understanding (MOU) was signed this year between the Landsat 7 Project and the EOS AM1 Project. The MOU specifies that the two systems will be flown in "loose formation" covering the same ground track 15-60 minutes apart. Significant benefits are expected from this arrangement including:
What the capabilities of Landsat 7, the spacecraft, instrument, and ground system, and the affiliation of the Landsat 7 and EOS AM1 missions amounts to is the expectation that better thematic mapper type data, better in terms of data quality and applicability, will be available faster and at less cost than at any previous time in the program's history. The potential information content of Landsat 7 data will be superior to that of its predecessors.
2.5 Program Management
As established by PL102-555 in October 1992, and amended by the National Space Technology Council/Presidential Decision Directive of May 5, 1994, the Landsat Program is managed by NASA, NOAA and the US Geological Survey. NASA is responsible for building and launching Landsat 7. NOAA will operate and maintain the satellite for the life of the system and manage ground system operations. The USGS will perform data capture, data processing and distribution functions on NOAA's behalf, and will store the data as part of the National Satellite Land Remote Sensing Data Archive (NSLRSDA).
The official launch date for Landsat 7 is May 15, 1998.
3. After Landsat 7
The path is obscure for Landsat-type remote sensing after Landsat 7. The ambiguity is caused by several factors including the requirements of the law, uncertainty about the direction Congress and the administration may choose to follow, the prospects for success of commercial remote sensing from space, the development of new technology, requirements for data, and the strength of the national will regarding technological leadership in earth remote sensing. The Landsat Program Management is seeking a strategy to guide the program through this morass and assure that public requirements for data are met.
3.1 Legislative Mandate
Any attempt to predict the future of the Landsat Program after Landsat 7 must begin with a review of the pertinent provisions of the 1992 law. The only change in the program of PL102-555 since enactment was the re-configuration of Landsat Program Management from NASA and DoD to NASA, NOAA and USGS. The law authorized the President to amend the management structure as deemed necessary, and President Clinton instituted the change in 1994 when DoD asked to withdraw from the program. All provisions of the law are being implemented
3.2 Title IV
The provisions in the 1992 law of particular relevance to the post Landsat 7 era are contained in Title IV - "Assessing Options for Successor Land Remote Sensing Systems" The law calls upon Landsat Program Management to report to Congress by October 1997 on options for a follow-on to Landsat 7. The report is required to include a "...full assessment of the advantages and disadvantages of 1) private sector funding and management of a program; 2) creating an international consortium for a follow-on system, 3) funding and management of the system by the US government, and; 4) build and operate a system through a cooperative effort between the US government and the private sector." The criteria for evaluating the options are also stated in the law as goals for the system. The follow-on to Landsat 7 should : 1) serve adequately the civilian, national security, commercial and foreign policy interests of the United States; 2) maintain data continuity with previous Landsats, and 3) include system enhancements that make the system cheaper to build and operate and more responsive to user data requirements.
Title IV contains one more section that is worth quoting, because it probably represents the current opinion of Congress on a Landsat 7 follow-on. Section (c) to Title IV reads, "Preference for Private Sector System - If a successor land remote sensing system to Landsat 7 can be funded and managed by the private sector while still achieving the goals [stated above] without jeopardizing the domestic, national security, and foreign policy interests of the United States, preference should be given to the development of such a system by the private sector without competition from the United States Government."
3.3 Actions to Date
The law presents a conundrum to Landsat Program Management. Continuity of Landsat-type data is defined as critical, something worth maintaining. There are few who dispute that point although what constitutes data continuity is a subject for discussion. Landsat 7 has a six year design life; therefore 2004 is the target date. A new system must be ready for launch that year. Construction must begin well prior to 2004. NASA is working to develop systems that can be designed, built and launched in 2-3 years. But, to date, there is no record of success for such systems on which to base future planning. It is only prudent to assume a longer lead time than 2-3 years will be required for a Landsat 7 follow-on.
The Landsat Program must make a recommendation by October of next year on how to maintain data continuity. However, to assure timely implementation of any option selected after October, 1997, actions must be taken now. NASA is therefore fulfilling the mandate of Title IV and, concurrently, following a course of action that will not preclude the option or options chosen.
3.3.1 Assessment of Options. To begin addressing the mandate of Title IV, NASA commissioned an assessment of options for "Landsat 8" based on the current state of land remote sensing from space and the stated plans for new commercial and government sponsored systems The report found that the primary demand for Landsat came from the global change science community. In the next five years, a number of systems will provide data similar to ETM+. That data may be more than adequate not only for the needs of the science community but for all users of Landsat data. The report recommended that the US government pursue purchase of data from the private sector to meet the need for Landsat-type data. If that option was not viable, the purchase of a commercial instrument for operation by the government combined with a commercially operated ground system and an international consortium to share data would likely be the most efficient way to proceed. Regardless of the option chosen, the report recommended that NASA return to its role of advanced technology development and retire from the responsibility of maintaining an operational remote sensing system.
The report was one element of a comprehensive strategy toward fulfilling the requirements of the law. Its recommendations have neither been adopted, nor dismissed. The viability of commercial systems has yet to be demonstrated. A data purchase has attractive features, but that type of program has no successful precedent at any scale, let alone the quantity of data a Landsat-type data purchase may entail. A crucial element in any data purchase will be the data policy, i.e., the rules for sale and redistribution of data after purchase. Who will have access to the data purchased by the government? What will it cost? How can it be used? The 1992 law found that commercialization of Landsat data was a failure because, in part, the cost of the data impeded distribution. What are the obligations of the US government to supply "low cost" data, to what elements of the user community, and at what level of data processing? Should the concept of COFUR, explicit in the Landsat 7 data policy, as required by law, be extended to the Landsat 7 follow -on? These issues will be addressed by any option selected for a post Landsat 7 system.
Finally, it remains far less than certain that the US government is prepared to retire its capability in Landsat type remote sensing after Landsat 7. TM data is still the data of choice of its type worldwide. The international ground stations now receiving Landsat 5 data seem eager to receive ETM+. Will the US walk away from this technology after Landsat 7? Is the US government prepared to rely exclusively on US commercial and non-US sources for Landsat-type data? These are crucial, and unresolved, questions.
Landsat Program Management has taken other actions to maintain flexibility to select an option or options after October 1997.
3.3.2 Landsat in EOS. The Landsat Program has been incorporated into the Earth Observing System. (EOS). Landsat-type data has been identified by the global change science community as crucial. It fills a resolution gap between the course resolution instruments such as AVHRR and MODIS and the very high resolution sensors the private sector is planning to operate. A Landsat type sensor was not included originally on the EOS platforms because it was assumed that such data would be made available through Landsat 6 and 7. However, to assure availability of Landsat-type data, a slot on the EOS-AM2 platform has been reserved for a Landsat-type sensor. AM2 is scheduled for launch in 2004.
3.3.3 New Millennium Program. The New Millennium Program was initiated this year and described in a paper presented in this symposium by the Landsat Project Scientist (Willimas, 1996). The program is designed to develop and test new technology for incorporation in future NASA mission instruments The first project approved under the program was Earth Observer -1 an instrument that may serve as the prototype for the follow-on to Landsat 7 on AM2 or on another platform. EO-1 will test the ability of the community to build, launch and operate a system in 2-3 years. It will be flown concurrently with Landsat 7 which will allow for comparison of data from the two systems. It will, in effect, supply the experimental instrument to compare with the operational instrument as was done with TM and MSS.
After almost 25 years, the Landsat Program remains as ill-defined as ever. Landsat 7 will be launched and will provide time to consider the future of the program. But, many of the features associated with "operational" satellite remote sensing programs are missing from Landsat. There is no back-up to Landsat 7 (ASTER is the nominal back-up, but ASTER was not intended or designed to meet a duty cycle similar to ETM+ nor is the ground system designed for,or capable of, processing and distributing ASTER products in quantities approaching that of Landsat 7.) There is no comprehensive plan in place to continue Landsat -type remote sensing after Landsat 7.
The uncertainty is due primarily to the changing nature of the earth observation from space. Between now and the launch of Landsat 7 many of the questions raised in this presentation will have clearer answers. Those answers are pre-requisite for development of a reasonable, coherent US policy toward acquisition and distribution of Landsat type data. Until those answers are available, Landsat Program Management is maintaining its options. It is likely that the US government will not invest another $800 million, the cost of Landsat 7, to build a stand alone Landsat 8.
It may not have to. A better understanding of the data requirements and the demonstrated capabilities of new systems may lead to the evolution of a program that provides the Landsat data user community with high quality data, in a timely manner and at reasonable cost. Unless and until instructed otherwise, acquisition and delivery of that data remains the primary objective of Landsat Program Management.