During the last few decades of the 20th century it became evident that tropical rainforests were endangered not only by road-building, timber-cutting, and other incursions of the bulldozer and saw, but by thousands of wildfires. Historically, relatively small fires have been set by slash-and-burn farmers trying to clear patches of jungle for farm land. But starting in 1997, fires in the worlds tropical forests from Brazil to Vietnam raged on a scale never recorded before. The causes of these huge conflagrations raised questions, because tropical rainforests rarely burn naturally.
In the wake of several haze-induced accidents and public health warnings in smoke-covered Indonesia, the need for a clear answer to these questions was given legal significance when President Suharto, under pressure from neighboring countries, passed a decree making it illegal to set forest fires. The politically connected timber industry had managed to direct most of the blame for forest fires on the small-scale, slash-and-burn farmers, but Indonesias rogue environment minister Sarwono Kusumaatmadja employed a relatively new intelligence-gathering technology to get a clear picture of the situation: he downloaded satellite images of burning Indonesian rainforests from a U.S. government website and compared them to timber concession maps. The satellite images confirmed that many of the blazes were being set in areas the timber companies wanted to clear for plantations. With the satellite evidence in hand, Kusumaatmadja got his government to revoke the licenses of 29 timber companies.
High Speed Intelligence
The environment ministers quick work on the rainforest issue is just one of many recent cases involving environmental questions in which satellite surveillance has been used to provide answers that might otherwise not have been known for years, if ever. The images captured by cameras circling high above the planet are proving effective not only because they scan far more extensively than ground observers can, but because they can be far faster than traditional information-gathering methods.
Pre-satellite studies of the oceans, for example, had to be done from boats, which can only reach a tiny fraction of the oceanic surface in any given month or year. And even after centuries of nautical exploration, most of the information scientists have gathered about winds, currents, and temperatures comes from the commercial trade routes of the North Atlantic between the United States and Europe. Satellites dont replace on-the-water research, as they cant collect water samples, but for some kinds of data collection they can do in minutes what might take boats centuries to do. Satellites can, in principle, watch the whole of the worlds oceans, providing almost immediate assessments of environmental conditions everywhere.
Similarly quick surveillance is available for many parts of the biosphere that are otherwise difficult to reachthe polar ice, dense forest interiors, and atmosphere. As a result, says Claire Parkinson of the U.S. National Aeronautics and Space Administration (NASA), theory and explanations no longer have a database restricted to areas and times where humans have physically [gone] and made observations or left instruments to record the measurements. The speed of environmental research has taken a quantum leap.
Speed isnt only a matter of technical capability, however. In practice, its also a matter of access. Spy satellites began circling the globe soon after Russias Sputnik went into orbit in 1957. But the information they relayed to Soviet and U.S. intelligence agencies was kept sequestered. The difference now is that satellites are increasingly being used for purposes other than espionage or military intelligence. The great majority are for telecommunications. However, more than 45many owned by governments, but a growing number of them privately ownedare being used for monitoring various phenomena on the ground, on the water, or in the atmosphere. In addition, more than 70 launches are planned during the next 15 years by civil space agencies and private companies. How these instruments are used, and by whom, will have enormous consequences for the world.
The Race Against Time
If incidents like the Indonesian forest fire intervention are any indication, environmental monitoring by orbiting cameras could play a critical role in reversing the global trends of deforestation and ecological collapse that now threaten the long-term viability of civilization. Denis Hayes, the former Worldwatch Institute researcher who is the chairman of Earth Day 2000 (see page 6) asked a few years ago, in a speech, How can we have won so many environmental battles, yet be so close to losing the war? Since then, we have edged still closer. Clearly, the number of battles being won is too small, and the time it takes to win them is too long.
Another way of posing Hayess now famous question might be to ask whether the processes of information-gathering and dissemination essential to changing human behavior can be speeded up enough to accelerate the environmental movement. Telecommunications satellites began providing part of the answer several decades ago by facilitating the formation of an active international environmental community that can mobilize quicklywhether to protest a dam on the Narmada River of India or to stop the use of genetically modified organisms in food production in Europe.
But while activism gained momentum, field work remained ominously slowbiologists slogging about in boots and rowboats, while the forces they were trying to understand raced over the Earth on the wings of global commerce, or ripped into it with the blades of industrial agriculture and resource extraction. However, satellite monitoring has begun to help researchers to more quickly assemble the data needed to bring decisive change. Remotely sensed images are contributing to critical areas of environmental research and management, including:
Weather: The first meteorological satellites were launched in the early 1960s, and quickly became a key part of the U.N. World Meteorological Organizations World Weather Watch. In addition to the satellite data, virtually all nations contribute surface measurements of temperature, precipitation, and wind to this program to aid weather prediction, which has enormous social and economic benefits. In recent years, optical sensors that collect data on sea surface temperature and radar sensors that estimate ocean height have proven useful in understanding and predicting El Nio events, which can damage fisheries and agriculture by bringing warmth and wetness to much of the west coasts of South and North America, and drought to Southeast Asia, Australia, and parts of Africa.
Climate: In the 1990s, researchers began to delve into satellite archives to study longer-term climate patterns. For instance, satellite images have helped reveal a decrease in snow in the Northern Hemisphere, a lengthening growing season in northern latitudes, and the breakup of major ice sheets. Radar sensors have been used to construct topographical maps of the ocean bottom, which in turn provide better understanding of the ocean currents, tides, and temperatures that affect climate. However, it was not until recently that space agencies began to design satellite systems dedicated specifically to climate research. In 1999, the United States launched Terra, which carries five different sensors for recording climatic variables such as radiative energy fluxes, clouds, water vapor, snow cover, land use, and the biological productivity of oceans. It is to be the first in a series of satellites that will create a consistent data set for at least 18 years.
Coastal boundary changes: Whether as a result of warmer temperatures contributing to sea-level rise or irrigation projects shrinking lakes, coastal configurations change over timesometimes dramatically. Scientists have compared declassified images from covert U.S. military satellites pointed at Antarctica in 1963 to recent images of the same regions, for example, to reveal changes in the continents ice cover. Other comparisons show the extent to which central Asias Aral Sea and Africas Lake Chad have diminished in size.
Habitat Protection: The destruction of habitats as a result of human expansion has been identified as the largest single cause of biodiversity loss. Satellite images have proved quite effective in revealing large-scale forest destruction, whether by fire or clearcutting, not only in Indonesia but in the Amazon and other biological hotspots. New, more detailed imagery may reveal small-scale habitat niches. In Australia, for example, the Australia Koala Foundation plans to use detailed satellite images to identify individual eucalyptus trees. Researchers will compare these images to field data to determine what this species of tree looks like from above, then use the information to more quickly map individual trees or groves than would be possible from the ground. In the oceans, the same satellite-generated maps of undersea topography and sea surface temperature used to study weather and climate can be used to track the upwellings of nutrient-rich water that help to sustain fisheries.
Environmental law enforcement: International organizations and national governments can use remote imaging to put more teeth in environmental laws and treaties. One of the leading fishing nations, Peru, is monitoring its coastal waters to prevent the kind of heavy overfishing that has so often caused fisheries to collapse. In Italy, the city of Ancona plans to buy satellite images to detect illegal waste dumps. Within the next decade, large-scale use of this technology could give urgently needed new effectiveness to such agreements as the Kyoto Protocol to the Climate Convention, the Biodiversity Convention, the Convention on Illegal Trade in Endangered Species (CITES), or the Law of the Sea.
Making Sense of Nonsense
Look closely at a small detail of a newspaper or magazine photoput it under a magnifying glassand it may make no sense. The dots dont form any recognizable image. But stand back and see the photo as a whole, and it snaps into focus. Satellite images do the same thing, only on a vastly larger scale. Bits of information that might make no meaningful pattern when seen from the ground may, when seen from many kilometers above, resolve themselves into startling pictures.
The first pictures from space were photographs made from film, by astronauts aboard the first manned flights to the moon in the 1960s. These photos, of a fragile blue planet suspended in the vast blackness of space, helped to inspire the nascent environmental movementone of them becoming the emblem of the first Earth Day in 1970.
In later surveillance from satellites, the imaging was digitized so that the data could be sent down in continuous streams and in much larger quantities than would be possible with film. Although one Russian satellite still uses regular camera film that is dropped to Earth in a canister and retrieved from the North Sea, most remote sensing satellites now use digital electronic sensors. The binary data they send down can be reconstructed into visual images by ground-based computers.
The amount of detail varies with the type of sensor (see table, page 21). For instance, an image of a 1,000 square kilometer tract of land obtained by a fairly low-resolution satellite such as AVHRR, which is used in continental and global studies of land and ocean, might contain 1,000 picture elementsor pixels (one piece of data per square kilometer). In contrast, an image of the same tract from the new, high-resolution Ikonos satellite would have 1 billion pixels (one per square meter). Between the broad perspective of satellites like AVHRR and the telescopic imaging of those like the new spy-quality satellite Ikonos, are medium-resolution sensors, such as those aboard the Landsat and SPOT satellites, in which one pixel represents a piece of land that is between 30 and 10 meters across (see back cover). However, the level of detail is also limited by the size of the medium on which an image is displayed. For instance, if an Ikonos image with 1 billion pixels were reproduced in a magazine image one-quarter the size of this page, it would be reduced to 350,000 pixels.
Different tasks require different levels of detail. The value of high resolution lies in its enabling the viewer to hone in on a much smaller piece of the ground and see it in a kind of detail that the lower resolution camera could not capture. For a larger area, a lower resolution would suffice to give the human eye and brain as clear a pattern as it can recognize. Whereas the wide coverage provided by lower-resolution satellites has proved useful in understanding large-scale natural features such as geologic formations and ocean circulation, very detailed imagery may be best able to reveal niche habitatssuch as the individual treetops that are home to the koalaand manmade structures, such as buildings, tanks, weapons, and refugee camps.
But satellite surveillance can do much more than provide huge volumes of sharp visual detail of the kind recorded by conventional optical cameras. The orbital industry also deploys a range of sensors that pick up information outside the range of the human eye, which can then be translated into visual form: Near-infrared emissions from the ground can be used to assess the health of plant growth, either in agriculture or in natural ecosystems, because healthy green vegetation reflects most of the near-infrared radiation it receives; Thermal radiation can reveal fires that would otherwise be obscured by smoke; Microwave emissions can provide information about soil moisture, wind speed, and rainfall over the oceans; Radarshort bursts of microwaves transmitted from the satellitecan penetrate the atmosphere in all conditions, and thus can see in the dark and through haze, clouds, or smoke. Radar sensors launched by European, Japanese, and Canadian agencies in the 1990s have been used mainly to detect changes in the freezing of sea ice in dark, northern latitudes, helping ships to navigate ice fields and steer clear of icebergs. Radar is what enabled satellites to map the ocean bottom, which would otherwise be obscured.
While satellites have the technical capability to monitor the Earths entire surfaceday or night, cloud-covered or clear, on the ground or underwaterthis doesnt mean we now have updated global maps of whatever we want. Aside from the World Weather Watch, there is no process for coordinating a worldwide, long-term time series of comparable data from Earth observations. Rather, individual scientists collect data to answer specific questions for their own projects. In recent years, national space administrations have teamed up with research funding agencies and two international research programs to support an Integrated Global Observing Strategy that would create a framework for uniting environmental observations. Researchers are now trying to demonstrate the viability of this approach with a suite of projects, including one on forests and another on oceans.
In addition, to make sense of remotely sensed data requires comparison with field observations. An important element of the weather programs success, for instance, is the multitude of observations from both sky and land. And satellite estimations of sea-surface temperature cant generate El Nio forecasts automatically, but must be combined with other data sources, including readings from a network of ocean buoys that monitor wind speed and a satellite altimeter than measures water height. Similarly, the radar scans used to make maps of the ocean floor are calibrated and augmented by sounding surveys conducted by ships. Even a task as straightforward as the location of eucalyptus trees for the Australia koala project requires initial field observations to confirm that the typical visual pattern being searched out from above is indeed that of the eucalyptus, and not of some other kind of tree.
Satellite imagery has become even more useful with the advent of geographic information systems (GIS), which allow users to combine satellite images with other data in a computer to create maps and model changes over time. In much the same way that old medical encyclopedias depict human anatomy, with transparencies of the skeleton, circulatory system, nervous system, and organs that can be laid over a picture of the body, a GIS stores multiple layers of geographically referenced information. The data layers might include satellite images, topography, political boundaries, rivers, highways, utility lines, sources of pollution, and wildlife habitat.
Maps that are stored in a GIS allow people to exploit the data storage capacity and calculating power of computers. Thus when geographically referenced data are entered into a GIS, the computer can be harnessed to look at changes over time, to identify relationships between different data layers, to change variables in order to ask what if questions, and to explore various alternatives for future action.
Because human perception can often identify patterns more easily on maps than in written text or numbers, maps can help people understand and analyze problems in ways that other types of information cannot. The Washington, D.C.-based World Resources Institute (WRI), for example, has used GIS to analyze threats to natural resources on a global scale. Researchers have combined ground and satellite data on forests with information about wilderness areas and roads to map the worlds remaining large, intact frontier forests and identify hot-spots of deforestation. A similar WRI study, investigating threats to coral reefs, assembled information from 14 global data sets, local studies of 800 sites, and scientific expertise to conclude that 58 percent of the worlds reefs are at risk from development.
With advances in computing power, some GIS software packages can now be run on desktop computers, allowing more people to take advantage of them. In fact, the number of people using GIS is swelling by roughly 20 percent each year, and the leading software company, ESRI, grew from fewer than 50,000 clients in 1990 to more than 220,000 in 1999.
A related technology spurring the market for geographic information is the Global Positioning Systema network of 24 navigation satellites operated by the U.S. Department of Defense. A GPS receiver on the ground uses signals from different satellites to triangulate position. (For security reasons, the Defense Department purposefully introduces a distortion into the signal so that the location is correct only to within 100 meters.) As GPS receivers have become miniaturized, their cost has come down. The technology is now built into some farm machines, cars, and laptop computers. Researchers can take air or water samples and feed the data directly into a GIS, with latitude and longitude coordinates supplied by the GPS receiver in their computers.
The relatively new field of precision agriculture demonstrates how satellite imagery, GIS, and GPS systems can all be used to show farmers precisely how their crops are growing. Conditions in every crop row can be monitored when a farmer walks into the fieldor when a satellite flies overand recorded for analysis in a GIS. During the growing season, satellite monitoring can track crop conditions, such as the amount of pest damage or water stress, and allow farmers to attend to affected areas. The central component of a precision agriculture operation is a yield monitor, which is a sensor in a harvesting combine that receives GPS coordinates. As the combine harvests a crop such as corn, the sensor records the quality and quantity of the harvest from each section of the field. This detailed information provides indicators about the soil quality and irrigation needs of different parts of the field, and allows the farmer to apply water, fertilizer, and pesticides more accurately the following season.
Meanwhile, the growth of the Internet is allowing satellite images and GIS data to be more easily distributed. In late 1997, Microsoft Corporation teamed up with the Russian space agency Sovinformsputnik and image providers such as Aerial Images, Inc. and the U.S. Geological Survey to create TerraServer, the first website to allow people to view, download, and purchase satellite images. Some satellite operators have begun to offer catalogs of their images on the Internet.
Whose Picture Is It?
Throughout history, people have fought for possession of pieces of the Earths land and water. It was not until the advent of Earth observation satellites that ownership of the images of those places was seriously debated. To legitimize its satellite program, the United States argued strongly that the light reflected off the oceans or mountains, like the air, should be in the public domaina claim now accepted by many other countries. But for the last two decades, the United States has also promoted the involvement of private U.S. companies in earth observation. Until the end of the Cold War, companies were reluctant to enter the satellite remote sensing business for fear of restrictions related to national security concerns. But in September 1999, a U.S.-based firm called Space Imaging launched the first high-resolution commercial satellite; this year, two other U.S. enterprises, Orbimage and EarthWatch, plan to launch similar instruments.
This trend raises questions about the tension between public and private interests in exploiting space. On the one hand, the widespread availability of detailed images means greater openness in human conduct. With satellite imagery, it is impossible to hide (or not find out about) such harmful or threatening activities as Chernobyl-scale nuclear accidents, or widespread forest clearing, or major troop movements. On the other hand, whether the information will be used to its full potential is up to governments and citizens.
There are obvious benefits to commercially available high-resolution images. Governments wary of revealing secrets have traditionally restricted the circulation of detailed satellite information. Now, images of the Earth that were once available to a select few intelligence agencies are accessible to anyone with a credit card. The information may be valuable to many non-military enterprisespublic utilities, transportation planners, telecommunications firms, foresters, and others who already rely on up-to-date maps for routine operations.
In the world of NGOs, the impact of high-resolution imagery may be most dramatic for groups that keep an eye on arms control agreements and government military activities. When one-meter black-and-white pictures hit the market, a well-endowed nongovernmental organization will be able to have pictures better than [those] the U.S. spy satellites took in 1972 at the time of the first strategic arms accord, writes Peter Zimmerman, a remote-sensing and arms control expert, in a 1999 Scientific American article.
Indeed, when Ikonos released imagery of a top-secret North Korean missile base last January, the Federation of American Scientists (FAS), a U.S.-based nonprofit group, published a controversial analysis of the images contradicting U.S. military claims that the site is one of the most serious missile threats facing the United States. (The potential threat from this base is a key argument for proponents of a multi-billion dollar missile shield, the construction of which would violate the anti-ballistic missile treaty.) Noting the absence of transportation links, paved roads, propellant storage, and staff housing, the FAS report found the facility incapable of supporting the extensive test program that would be needed to fully develop a reliable missile system.
In addition, the private entrants in the satellite remote sensing business may spur the whole industry to become more accessible. Already, the new companies are beginning to seek partnerships with government imagery providers, so that customers are able to go to one place to buy a number of different types of images. For instance, Orbimage, which is scheduled to launch its first high resolution satellite this year, has made agreements to sell medium resolution images from the French SPOT series and radar images from the Canadian Radarsat. Such arrangements may make it easier for people to find and use imagery.
But private ownership of satellite datamaking it available only at a price that not all beneficiaries can pay, or protecting it with copyright agreementscould also cause serious impediments to reversing ecological decline:
One of the most important potential applications of remote sensing could be its use by non-governmental public interest groups, which provide a critical counterweight to the government and corporate sectors. NGOs that monitor arms control or humanitarian emergencies, for example, can use satellite data to pressure governments to live up to international nonproliferation agreements and foreign aid commitments. But a group that buys images from a private satellite company in order to publicize a humanitarian disaster or environmental threat coulddepending on copyright laws and restrictionsfind itself prohibited from posting the images on its website or distributing them to the media. While low-resolution imagery from government sources is easily shared, citizen groups and governments should quicky set a precedent for sharing information from high-resolution imagery.
Only a few, very large agricultural operations can afford the high-tech equipment required to practice precision agriculture. Because farmers with small plots of land can simply walk into their fields for an assessment, the technology will remain most useful to larger operations. So although this tool could improve the way large-scale agriculture is practiced in the short term, it could also delay the long-term transition to more sustainable agricultural practices, which tend to require smaller-scale farms.
Large companies looking for places to extract oil, minerals, or biological resources could purchase detailed satellite data and gain unfair advantage over cash-strapped developing nations in which the resources are located, whose governments cannot afford satellite imagery, GIS software, and technical support staff and systems needed to make and maintain such maps.
Finally, along with the question of who will own the technology, there is the related question of whether there is significant risk of its being badly misused. A few decades ago, the advent of commercially available high-resolution images would likely have been met with great alarm, as a manifestation of the Big-Brother-is-watching-you fear that pervaded the Cold War years. That fear may have receded, but what remains is a conundrum that has haunted every powerful new technology, from steel blades to genetic engineering. Could access to detailed images of their enemies cause belligerent nations to become more dangerous than they already are?
In any case, theres no turning back now. Many of the remote sensing satellites now scanning the Earth and scheduled for launch this year will be in orbit until long after the basic decisions affecting the planets long-term environmental futureand perhaps the future of civilizationhave been made. Ultimately, an educated global citizenry will be needed to make use of the flood of data being unleashed from both publicly and privately owned satellites. Policy analyst Ann Florini of the Carnegie Endowment writes: With states, international organizations, and corporations all prodding one another to release ever more information, civil society can take that information, analyze and compile it, and disseminate it to networks of citizen groups and consumer organizations.
If the Internet has given the worlds technological infrastructure a new nervous system, the Earth-observation satellites are giving it a new set of eyes. In precarious times, that could be useful. U.S. Vice President Al Gore, who understands the potential of remote sensing, has called for the completion of a Digital Earth, a 1-meter resolution map of the world that would be widely accessible. According to Brian Soliday of Space Imaging, the Ikonos instrument alone might be able to assemble a cloud-free map of the world at 1-meter resolution within four to five years. Thats about as long as it has taken to do the vaunted Human Genome map, and this map would be much bigger. Arguably, because it covers not only us humans but also the biological and climatic systems in which our genome evolved and must forever continue to depend, it could also be at least as valuable.
Molly OMeara Sheehan is a research associate at the Worldwatch Institute.