High-resolution satellite imagery, previously restricted to intelligence organizations, is about to make a commercial debut. In the next two years, U.S. companies plan to launch satellites carrying sensors capable of recording and transmitting images of the earth's surface with one-meter resolution. Fully processed images will be precise enough to meet U.S. accuracy standards for 1:2400 scale mapping (1.5-meter positional accuracy) -- detailed enough to spot a motorcycle from 400 miles out in space.
Before the end of the decade, providers will downlink, process, and distribute black and white, color, infrared and stereo pairs to customers via the Internet, all within hours or days of an order, depending on the product requested. Both special and off-the-shelf imagery will be compatible with GIS, CAD and desktop mapping systems, and will run on PCs having sufficient memory and disk space to handle map files.
Customers will have the option of buying fully processed images ready for GIS/CAD integration or processing the raw data on their own. Knowledgeable buyers with established accounts will shop for satellite imagery at vendors' Web sites by simply entering the coordinates of an area and providing accompanying information.
Before the end of the decade, affordable, high-resolution, space-based imagery will undoubtedly find application in nearly all levels of city and county government, as well as across the entire spectrum of U.S. commerce. The combination of domestic and foreign competition may put the cost of high-resolution imagery well below that of aerial photography.
Until recently, the same circumstances that blocked commercial development of space-based imagery in the United States effectively limited the image resolution of satellites owned by foreign companies. As of this writing no image-recording satellites are owned by U.S. companies, nor can any commercially available satellite imagery match the spatial resolution and positional accuracy of aerial photography. All that, however, is soon to change.
Spot Image, S.A., of Toulouse, France, began producing 10-meter resolution imagery from its own satellites as early as 1986, and today is a major supplier in this market. Other foreign companies quickly followed -- the Japanese JERS satellite, providing 20-meter resolution; Europe's remote sensing ERS-1 and ERS-2 satellites offering 10-meter resolution; and India's IRS-1C, with 5-meter resolution -- all with black & white imagery.
With even Russia selling 5- and 8-meter resolution from their KFA-1000 and MK-4 satellites, respectively -- and on very rare occasions, letting go of 2-meter military images -- security concerns began falling through the cracks. In 1992, Congress passed the Land Remote Sensing Act, allowing private companies to launch satellites with sensors capable of producing 3-meter resolution. In 1994, the Clinton administration bumped that to 1-meter resolution.
Before looking at currently available satellite imagery, it might help to review some concepts and typical applications associated with this technology.
Specific wavelengths of light recorded by orbiting sensors not only provide different levels of detail but different information. Panchromatic, or pan, (black and white) has a much higher resolution than multispectral color imagery. Pan sensors capture a broad spectrum of light in a single measurement in which each pixel (smallest picture element that can be individually addressed) is assigned a specific gray-scale value. In 1-meter resolution imagery, each pixel represents an area of about three feet.
Multispectral imagery is made up of three or more bands -- blue, green, red, infrared, and parts of the thermal spectrum. Each band is separately recorded, processed and integrated into a composite color image. Since pan delivers considerably higher resolution, it generally costs more than color.
Ten-meter pan imagery, presently a commonly available product, has applications in regional, county and urban planning; updating road maps, looking at subdivisions, individual houses or buildings, mapping street centerlines and evaluating site proposals. To make map information more understandable as well as attractive, pan imagery