2
Models for Providing Remote Sensing Data for Scientific Research
In the early 1990s, the federal government was the principal provider of remote sensing data to the scientific community,1 but it had already begun to encourage a private sector role in supplying the data. The Land Remote Sensing Policy Act of 1992 directed Landsat Program Management (NASA and DOD) to assess four approaches for developing the Landsat follow-on program: (1) a government effort, (2) a commercial-private sector effort, (3) a government-private sector cooperative effort, and (4) an international consortium.2
In the decade since the 1992 act, the federal government and private sector have explored the first three of the four approaches; an international consortium has not yet been attempted. Although these four models were suggested specifically as options for the Landsat follow-on, they also provide a useful framework for considering the institutional arrangements for providing new remote sensing data for scientific research. Three of these approaches—government, commercial, and public-private—were discussed at the steering committee’s March 2001 workshop.3 This chapter outlines the four models.
MODEL 1: THE GOVERNMENT AS DATA PROVIDER FOR SCIENCE
The first model—with the government as data provider—is the traditional approach. It assumes that federal science agencies and operational agencies will continue to launch and operate satellites that provide data for scientific research. Examples are NASA’s Terra satellite and NOAA’s Advanced Very High Resolution Radiometer (AVHRR). Examples still in development are NASA’s Aura mission and the NOAA-DOD-NASA National Polar-orbiting Operational Environmental Satellite System (NPOESS).4 The operating assumption behind this approach is that the data are primarily intended for public purposes, such as research and government operations, rather than for for-profit uses in the commercial marketplace. In cases where the government is the data provider of choice for science, NASA and NOAA (and their partners) will continue to dominate in the production of the data.
Landsat 7 is an example of the government-provider approach. Although developed by NASA, it is currently operated by the U.S. Geological Survey (USGS). USGS also archives and disseminates Landsat 7 data, which are available to all users at the cost of reproduction.
MODEL 2: THE PRIVATE SECTOR AS DATA PROVIDER FOR SCIENCE
The model involving the private sector as data provider is a more experimental approach, made possible by recent legislation.5 In this model, the private sector finances, builds, launches, and operates a satellite, making data available on a commercial basis for multiple purposes, including research. The government may be a user of the data, but it does not specify data requirements. At present, the best and possibly only examples of completely private sector Earth observation satellites are IKONOS and QuickBird, although there are 12 fully private sector systems now planned for launch during the next 5 years.6
Transactions between scientists and private sector satellite providers may occur on an individual basis (e.g., scientists may use funds from research grants
4 |
The National Performance Review and Presidential Decision Directive of 1994 directed the DOD (Air Force) and U.S. Department of Commerce (NOAA) to create a converged weather satellite program that would meet U.S. civil and national security requirements and fulfill international obligations. NPOESS is the converged system. |
5 |
The Land Remote Sensing Policy Act of 1992 (Public Law 102-555) allows for the Secretary of Commerce to issue licenses to private sector parties to operate private remote sensing systems. Presidential Decision Directive 23 lays out the terms for licensing and operation of private remote sensing systems. |
6 |
William E. Stoney, “Summary of Land Imaging Satellites Planned to be Operational by 2006,” Mitretek Systems, March 2, 2001, p. 3. Available online at <http://www.asprs.org/asprs/news>. Accessed September 25, 2001. |
to procure satellite data from aerial remote sensing firms or commercial satellite remote sensing firms); however, many scientists may not have the funding to purchase research data from the private sector. NASA grantees can submit a competitive proposal and, if selected, receive commercial remote sensing data at no cost through the Science Data Buy (SDB) program, which is managed out of NASA’s Stennis Space Center (see Chapter 3). If funding for the SDB program does not continue, these researchers may no longer be able to obtain commercial data, as they may lack the resources in their grants to purchase the data directly at market prices. However, as the number of private remote sensing satellite data collectors increases, market forces and competitive pricing could make commercial data more affordable to scientists. As the use of commercial remote sensing data for scientific research evolves, several issues must be considered—including data management, data processing, long-term archiving, and intellectual property and data access—if privately collected data are to meet the requirements of and the broader needs for scientific research. (These issues are discussed in Chapter 4.)
MODEL 3: PUBLIC-PRIVATE SECTOR RELATIONSHIPS FOR PROVIDING SCIENCE DATA
The public-private sector approach can take several forms. An early example was the privatization of the government’s Landsat program in the mid-1980s. Landsats 1, 2, 3, and 4 had been developed in the 1970s and 1980s and were initially operated by the federal government as an experimental system. According to a report of the Space Studies Board, Office of Management and Budget (OMB) approval of Landsat assumed that markets would grow to provide funding for all future systems.7 However, government policy makers saw the slow initial growth of the market as a programmatic failure. The OMB had supported an operational weather satellite system, but was reluctant to support Landsat as an operational program. The budget situation for Landsat had been uncertain from the very beginning. In the late 1960s, just as Landsat reached the development stages, space budgets were beginning to incur serious cutbacks. There was a need to “sell” NASA, the Congress, and the administration on the future growth of a private market, which was expected to underwrite the cost of the government’s data needs. A long debate ensued over the question of whether to maintain Landsat as a public program or transfer it to the private sector.8 The debate culminated in 1984 with the administration’s decision under the Land Remote-
Sensing Commercialization Act of 1984 (Public Law 98-365) to transfer Landsat to the private sector9 to operate and sell the data on the commercial market.
A series of steps led to the selection of the Earth Observation Satellite Company (EOSAT) as the private operator for Landsat, and in October 1985, the company took charge of the Landsat system (see Table 1.1 in Chapter 1). Thus, the government role changed midstream in the program from being developer, operator and distributor of Landsat data to being data purchaser, although the EROS Data Center continued to distribute data for the commercial operator until 1989. As a result of new ground processing technology, the data volume of Landsat digital tapes increased sevenfold. In turn, the government increased the price of a Landsat satellite scene seven times, to $4,400, in preparation for the transfer of the Landsat program to the private sector.10 Both NASA and EOSAT raised prices for photographic images of Landsat scenes cumulatively by 1,000 to 2,000 percent over the period from 1980 to 1990. Consequently, sales of the data to users, including the academic community, dropped significantly.11
During the 1990s, the U.S. government in effect tested different types of public-private sector partnerships or hybrid approaches to obtaining data expressly for scientific research. The Sea-viewing Wide Field-of-view Sensor (SeaWiFS) and NASA’s Science Data Buy (SDB) are examples of this approach. In both cases the government agreed to purchase data for scientific research in advance of its use. There was also a proposal from the Jet Propulsion Laboratory in 1996 to partner with industry in developing a small synthetic aperature radar program (called LightSAR) for both scientific and commercial applications.12 However, potential industrial partners did not believe that a viable market existed for data from a “system that focused primarily on meeting the NASA science requirements,” and they declined to participate.13 In addition, a proposed remote sensing venture between the Office of Naval Research and the private sector for a hyperspectral imager, the Naval EarthMap Observer (NEMO) satellite, faces challenges resulting from the commercial partner’s difficulties in attracting
necessary private sector investment. Private industry is reluctant to fund the system because it questions whether a commercial market exists for hyperspectral data.14
For SeaWiFS, the relationship between the government and the private sector was closer to the traditional approach of government supplier: NASA scientists specified the data requirements and were heavily involved in developing specifications for the instrument. Though the private vendor owned and operated the satellite and could sell the data commercially, the guaranteed purchase of data by the government enabled the company to obtain the private sector capital it needed to develop the system.
Unlike SeaWiFS, which was developed to meet science data needs, the SDB program acquires data from private sector firms—including both operational satellite and aerial remote sensing firms whose primary market is commercial and applied users, rather than researchers. For this program, NASA guarantees that it will purchase data at a specific cost for purposes of scientific research. The agency serves as middleman between scientists and private sector data suppliers. In addition, NASA provides verification and validation of all data distributed through the program.
Another variation on the public-private sector approach is exemplified by the case of the French remote sensing satellite, SPOT. Under this arrangement, initially conceived and developed as a public-private partnership, the French space agency supports the research and development of the satellites, and a quasi-private company, Spot Image, sells the data commercially. Spot Image includes among its investors government space agencies of France, Sweden, and Belgium, as well as private companies. Similar to the French system is Canada’s Radarsat 1, which was funded by Canada and developed by the Canadian Space Agency with contributions from the United States.15 A private company, Radarsat International (RSI), markets, processes, and distributes Radarsat 1 data.16,17
MODEL 4: INTERNATIONAL CONSORTIUM FOR PRODUCING SCIENCE DATA
The use of an international consortium for producing remote sensing data for scientific research was not discussed at the March 2001 workshop. This approach has been explored in other reports,18 and some organizations are currently using or working toward this model. As an example, the European Meteorological Satellite Organization (EUMETSAT) is an intergovernmental, member-funded organization of European states that launches, operates, and delivers meteorological data to end users and monitors the climate.19 In addition, several international remote sensing organizations use the consortium approach to coordinate remote sensing systems and data distribution, though they do not own the satellites jointly. An example is the World Meteorological Organization’s World Weather Watch, which coordinates the meteorological satellites of several countries and delivers unprocessed global weather data. (This model, which falls outside the committee’s charge, is not explored in this report.)
A number of reports from the National Research Council have focused on the role of government programs in providing remote sensing data for scientific research (Model 1). These include Global Environmental Change: Research Pathways for the Next Decade20 and several report of the Space Studies Board.21 The focus of the steering committee’s discussions and of the workshop was public-private sector partnerships (Model 3). The discussion in Chapter 3, “Toward Successful Public-Private Cooperation,” concentrates on the two most fully developed examples of such partnerships, SeaWiFS and SDB, and explores how they work, what their strengths and weaknesses are, and how effective they are in meeting scientific data needs. Although the two programs have been operating for only a few years, it is not too soon to highlight what is working well, to identify what needs improvement, and to call attention to issues that are not being adequately addressed.
18 |
See, for example, Office of Technology Assessment, U.S. Congress, Civilian Satellite Remote Sensing: A Strategic Approach, Washington, D.C., U.S. Government Printing Office, September 1994, pp. 101-128. |
19 |
For additional information, see the European Meteorological Satellite Organization Web site at <www.eumetsat.de/en/area1/topic1.html>. Accessed September 25, 2001. |
20 |
National Research Council, Global Environmental Change: Research Pathways for the Next Decade, Washington, D.C., National Academy Press, 1999. |
21 |
See reports of the Space Studies Board, Committee on Earth Science of the National Research Council: Issues in the Integration of Research and Operational Satellite Systems for Climate Research: Science and Design (2000); Issues in the Integration of Research and Operational Satellite Systems for Climate Research: Implementation (2000); The Role of Small Satellites in NASA and NOAA Earth Observation Programs (2000); and Development and Application of Small Spaceborne Synthetic Aperture Radars (1998), all published by the National Academy Press, Washington, D.C. |