Nearly 10 years ago, NASA - the U.S. National Aeronautics and Space Administration -launched the first in a series of unmanned space satellites designed to study the surface of the earth by a process called "remote sensing" and to provide geographers, geologists, cartographers and other environmental specialists with a unique perspective on the earth's surface and resources.
Remote sensing - a term coined by a geographer in the office of U.S. Naval Research - means, in effect, the accumulation of information from a distance: spotting a sail on the horizon through a telescope or detecting the approach of bombers on radar. With regard to the unmanned satellite program - the LANDSAT program - remote sensing refers to the process by which highly sophisticated instruments called "multi spectral scanners" record a unique kind of "image" of the earth's surface and, with the help of computers, extract invaluable information from the image about minerals, forests, agriculture, water, population and ecology on earth. So far the satellites have sent more than 300,000 "images".
Like most of the great technological innovations, remote sensing is the end product of years of development, but it didn't really get underway until 1838 when Daguerre worked out a way to take a photograph on a silver-coated copper plate. This was the Daguerreotype, a cornerstone of photography. Its development thereafter, however, was swift, as imaginative applications spurred progress in photographic processes and related technology. Soon afterwards, for example, cameras mounted on kites and balloons were pressed into the service of topography in France, in exploration of the American West, in 1850, and in military observation during the Civil War.
Later, sensitized emulsions replaced Daguerre's cumbersome plates, and other improvements came, one after the other: multi-layered color film, motion pictures, haze penetration reconnaissance film, pnotogrammetry techniques and a multitude of lenses and electronic accessories. The most recent advance - which came in conjunction with the perfection of space launch techniques and the introduction of special satellites in the 1960s - is the use of "multispectral images" produced by "multispectral scanners."
Modern developments in remote sensing from space really started with the photographs taken from V-2 rockets sent up by American scientists in 1946, but the next important step was not taken until 1965 when the Gemini III crew brought back 25 color photographs of the earth, the first of 1,000 photos taken by 10 Gemini missions and two Apollo missions and evaluated as "acceptable earth-looking images." They were taken between March 23,1965, and March 13,1969.
Simultaneously, NASA experts were moving toward improved meteorological satellite systems -TIROS and NIMBUS - to provide images of cloud patterns and atmospheric phenomena; eventually they developed unmanned satellites specifically designed to collect surface data on a systematic basis: the LANDSAT satellites, a series of three satellites, the first of which went up in July, 1972, placed in orbits precisely calculated to provide the optimum elevations for gathering and recording information about the earth's terrain.
Moving along near-polar, circular orbits at 24,000 miles an hour, 920 kilometers high (571 miles), these one-ton satellites take 103 minutes to orbit the earth - for a total of 14 orbits every day - and, since the satellites follow the sun, record images of specific locations at the same time every 18 days.
With regard to remote sensing, however, the satellite is just a platform. What is vital in the process is the sophisticated electronic sensor called the "multispectral scanner," the "eyes" of LANDSAT. The scanner measures the sunlight reflected off the earth's surface and can "see" further than the human eye in terms of scope. It picks up, for example, the reflected energy that the human eye can see, as well as other parts of the "electromagnetic spectrum." The solar part of this spectrum includes ranges of energy known as the "reflected" or "near infra-red" region - which lie beyond human vision.
On the other hand, the scanner cannot quite cope with the color blue - and this creates a problem. Though energy from the sun is perceived by the human eye as white light, the light actually consists of colors that are dominated by blue, green and red hues, as experiments with a prism or the sight of a rainbow demonstrate. The atmosphere, however, tends to scatter the blues from the energy - thus creating the blue sky – and this, in remote sensing, is a problem: because the atmosphere would reflect a disproportionate amount of blue back to the sensor, the scanner would be unable to "see" the terrain clearly. Thus the blue band was not included in the design of the multispectral scanner.
Another problem has to do with the infra-red region - the invisible radiation of the sun which is reflected back to the satellite's "eyes." The infra-red bands - Infra-red 1 and Infra-red 2 - have no color and so cannot be photographically expressed in normal colors. But they do have interesting and important characteristics that are valuable in assessing the data that images from space can provide; vegetation, if healthy, tends to reflect the "invisible light" while water tends to absorb it. Indeed, if human eyes could see this reflection in terms of black and white, the green vegetation might appear white and the water might appear black.
What the LANDSAT analysts are left with, therefore, are four bands: normal red and green and two invisible infra-red bands, all translated into numerical degrees of brightness and then transmitted from the satellite to receiving stations on earth where LANDSAT specialists create an image that can be subsequently transferred to, and stored on, film at the ground processing center.
This is done by using a computer to sort out assigned digital numbers relating to variations in the reflected brightness of features - mountains, sea or desert - and then create an image from that information, a system that, in fact, is as interesting as the launch of the satellite itself.
Essentially, this is what happens: the numerical data - or data in digital form - are stored on what are called "computer compatible tapes" (CCT’s) and the tapes are manipulated by the computer to "enhance" features of the "image." This is done in various ways by a' color additive system" resulting in images in which "false colors" are arbitrarily assigned to the various bands - and have nothing whatever to do with the real or visible color perceptible to the naked eye. Reds and pinks, for example, represent the infra-red energy bouncing off vegetation and growing crops; blues and greens represent urban areas, bare soil or plowed cropland; dark, uniform tones represent relatively clear water; light blues represent shallow water; dark, textured tones represent low-reflectance rocks; yellows and light tones often represent desert surfaces.
These variations in reflectivity -along with the individual characteristics of the scanner and the altitude of the orbit - also determine the amount of detail that can be seen on LANDSAT scenes. Each scene, for example, contains 7.6 million "pixels" - or picture elements - and covers a total of about 3.33 million hectares (8.2million acres). And each picture element represents an "average" of the reflectance of all objects in the area covered in the scene. As a result, only the larger features of the terrain are seen; gullies, gardens, pathways, individual houses or dwellings, dispersed settlements, small streams and small roads cannot be discerned; but urban areas, agricultural fields, major rivers, geological structures, forests and major transportation arteries can be detected - and their situation interpreted with respect to other information within the scene.
In the future, earth-oriented satellites and remote sensing systems will, undoubtedly, provide more detailed views of man's environment and will aid in improving his knowledge in various scientific fields. If, for example, the "pixel" size can be reduced - to say, 10 to 20 meters (32 to 64 feet) - the information in each scene will be substantially increased. Spectral expansion - via thermal infra-red channels that measure emitted infra-red energy from the earth - is also likely and would provide temperature information. Lastly, the LANDSAT program's fourth remote sensing satellite is expected to include a new system called a "thematic mapper."
Meanwhile, other countries are getting into remote sensing. In 1984, France will send up a system called SPOT, aboard an Ariane launch vehicle from French Guiana, to provide high quality, multispectral and panchromatic bands. And in Japan efforts are focused on the launch of MOS, another satellite with great potential.
Behind this surge of interest is the realization of the enormous benefits to be derived from space-imaging systems. Already, for example, ground stations in Brazil, Sweden, Italy, Japan, Argentina, Australia and India are receiving data that will eventually be invaluable in assessing mineral wealth, weather, the environment, and in planning - or controlling - urban growth.
In some areas, sensors and satellites have special value. Over regions such as the Netherlands or the British Isles, moisture and clouds interfere with remote sensing, but over arid lands -such as Egypt, the Arabian Peninsula and the Arabian Gulf countries—the imaging produced is extraordinary in terms or beauty.
Saudi Arabia, for example, though an exceptionally arid country, appears LANDSAT images as a series of marked contrasts: the rugged, rocky area of Makkah (Mecca); the broken coast of the Red Sea running north to the Sinai and Suez; the interior region of plateaus and sand deserts; the extensive plains of the eastern region, monotonously flat and masked by sheets of sand and beds of clay and marl; and on the eastern edge a giant hidden basin which yields oil and gas; and, finally, the warm waters of the Arabian Gulf, with coral reefs beneath the surface.
The imaging, however, is useful as well as beautiful. It helps in the search for water - the most critical resource for community and regional development in arid regions - as well as for minerals.
Characteristically, most water is found in aquifers in either porous sediments or greatly fractured bedrock. In the former - such as a sandstone bed -water is held as in a sponge, and in the latter, water is found in the cracks of otherwise dense rock. Geologic structures likely to yield water may be the fan of coarse sediments flanking mountain ranges - which reach to the clouds for their moisture - or highly fractured rock masses, as in the Makkah region shown on pages 18-19, and geohydrologists, from LANDSAT scenes, and such environmental indicators as vegetative growth -shown in red -plus traces of old stream channels can often pinpoint where test wells should be sunk.
The significance of this LANDSAT application is that aquifiers may be mapped on a national scale, thus providing the key to regional resource developments involving irrigation and water supply, the two most important water needs in the Middle East. Since LANDSAT scenes cover about 34,000 square kilometers (13,000 square miles), the regional magnitude of many geologic features can be more readily comprehended; aquifiers often extend over vast distances and neither ground views nor normal aerial photography, can fully encompass a geological unit of hydrologic significance or clearly evaluate its potential.
The same is true with regard to mineral resources, since geologic units of potential importance stretch for long distances. LANDSAT provides the broad view- the big picture - which is vital to recognizing and mapping potential sources. Indeed, the and landscapes of the Middle East are often a geologist's paradise; since no vegetative cover masks the surface, basic bedrock characteristics can be seen where exposures are not covered over by shifting sands or water-deposited sediments. LANDSAT images, moreover, not only reveal regional characteristics, but also provide information about specific types of rock because of the different reflective properties of rocks which the satellite picks up as green, red and infra-red light. Unique combinations of these reflectances provide "signatures" which are then attributed to specific kinds of rock.
The formation of valuable metalliferous ores - such as copper-and precious minerals is often associated with ancient structural features subdued long ago by the ravages of erosion. Faults, or zones, along the boundaries of crustal movement are expressed as major lineations on LANDSAT scenes; with the broad view provided by LANDSAT, these features are readily traced - often beyond their previously known limits.
Mineralization may also be associated with intrusions of molten material into previously existing crustal material, the slow cooling beneath the surface bringing about crystallization - which subsequent erosion exposes. On LANDSAT scenes, these often appear as near-circular "anomalies" contrasting with the surrounding terrain and may be so large as to defy comprehension at ground or aircraft levels of observation.
Spectral (color) signatures of the bedrock, traces of faults and other geologic lineations - such as joints, structural trend lines and crust displacement - the presence or absence of vegetation and topographic features are all incorporated into the geologic analysis of LANDSAT data and geoscientists must use them all to develop inventories and maps of the earth's resources.
In Saudi Arabia, such mapping programs have been underway since 1975. The Deputy Ministry for Mineral Resources, for example, has been operating a remote sensing applications facility since 1977, producing what are called "mosaics" from satellite-sensor images for use as base maps.
So far, three different series of these mosaic base maps have been started. Working with the U.S. Geological Survey and ERIM, the ministry has been working on l:250,000-scale quadrangle maps and has recently completed a l:2,000,000-scale mosaic of the entire peninsula composed of 256 images. Also finished is a 1:1,000,000,000-scale schematic soil map, the first ever attempt to produce a soil map for an entire country using LANDSAT data. And this year a l:2,000,000-scale geographic map of the peninsula will be finished, possibly by this spring and will show topographical detail never represented on comparable maps before.
Sensor images have also been used to track the diffusion in the Red Sea of 15,000 tons of mud pumped from the sea bottom - and back - as part of a mineral extraction experiment.
In sum, sensors and satellites show Saudi Arabia and the whole Middle East more graphically, more beautifully - and more usefully - than any system ever has before.
Aulis Hind, now director of the University of Vermont Remote Sensing Laboratory, and associate professor, won her Ph.D in geography from the University of Wisconsin and for 15 years has worked with NASA, the U.S. Army and the University of Vermont interpreting digitalized satellite photos.
