In a work table beneath the skylights of Douglas Way's office in a renovated Back Bay office building are two sheets of polyethylene that look like reproductions of abstract paintings.

One of them depicts the terrain where a 3,000-acre ski resort community is to be developed southeast of Pittsburgh. The other depicts most of the sultanate of Oman, where the government is trying to expand agricultural production.

Through an infrared process, the reproductions show vegetation in red. The image of Pennsylvania countryside appears almost all red; in bleak, rocky Oman, at the southern end of the Arabian peninsula, a few pinpricks of red locate oases and, through them, potential sources of subterranean water.

The representations aren't exactly photographs. They are reconstructions in photographic form of computerized data sent to Earth by orbiting satellites in space. But Way is no space scientist. He is a landscape architect, a member of the SWA Landscape Architects' group, one of the nation's biggest, and he uses the satellite data to plan large-scale land development and resource management for terrestrial clients.

Way is something of a pioneer in the application of photogrammetry--terrain analysis based on interpretation of aerial images--to land development, a technique that is apparently revolutionizing large-scale land planning.

There is nothing new about the idea of remote analysis of terrain; it has been going on since the first man climbed a hill to see into the distance. The first aerial photograph was taken from a balloon over Paris in 1858. What is happening now, according to experts in the field, is that satellite data make it possible to apply known techniques to larger and larger areas, and more geologists, landscape architects and agronomists are learning to do so. Their work represents the application to the civilian economy of information-gathering methods long used by the military.

"When I was in college in the 1960s," Way said, "there were maybe only 10 courses on photo interpretation in the whole country. Now there are over 500 college-level courses in remote sensing, but there are few sophisticated users of the data."

Way now teaches the subject at Harvard, where he is an associate professor in the Graduate School of Design, and at the Interior Department's Earth Resources Observation Systems (EROS) data center in Sioux Falls, S.D. EROS, one of several public sources of satellite data, makes information obtained by the Landsat satellites available to anyone who wants it.

"So much has changed in the past 3 1/2 years. There's a revolution going on," said William Campbell, who until recently ran a program at the Goddard Space Flight Center to instruct local government officials in the use of Landsat data for land planning. "In 10 years all data dealing with the Earth's surface will be in this format. We can provide billions of bits of information about the Earth's surface in an hour."

"This is all very new," said Edward H. Able Jr., executive director of the American Society of Landscape Architects. "We are just preparing a technical publication on it now. The field has just exploded--large scale land planning and land use analysis have taken on a whole new dimension. It's the same technique, but applied to larger and larger pieces of land."

"It's revolutionary for landscape architects to get involved in this," said Kalvin Platt, president of SWA. "People still think of us as gardeners," when much of their work now consists of large-scale land analysis and environmental planning in advance of development.

Platt and Way used the Pennsylvania project, to be called Nemacolin Woods, to illustrate the uses they can make of remote-sensing data.

Without setting foot on the land, they can identify vegetation patterns, stream beds, underground water supplies and seepage areas. They can determine the depth of the water table and the depth of the soil over bedrock, and calculate whether the land will permit the installation of septic tanks or require sewers. They can even distinguish between evergreen and deciduous trees.

"We can see where all the red flags are," Way said. "We want to stay out of the blue areas, that's underground water. It shows you water you can't see. It shows where to put a lake or a recreation area, where is the best site for houses."

"Eventually you still have to get out on the site and get your feet dirty," Platt said, "but on a site like this, with thick forest cover, you can't really see anything just by walking around."

Platt said his firm now performs such surveys for clients who are contemplating land purchase and development. The information lets the prospective buyer know in advance what kind of development the land will support, and how much. The client then can decide whether or not to proceed with the acquisition, Platt said.

According to Charles Kilpack, a landscape architect who is writing a technical manual on remote sensing, satellites currently in orbit provide photographic "resolution" of land areas as small as 1.1 acres. A new satellite, designated Landsat D, which is to be launched in July, will reduce that to half an acre, he said.

The sharpness of resolution currently available, he said, is not adequate for precision cartography, but he said it is possible to obtain "finer classification" by working directly from the computer-tape data, instead of from the photographic re-creations.

Way said the 1.1-acre resolution enables him and his associates at SWA to use the material for "anything from a natural resource survey of an entire country to specific site planning for 100 town houses." The scope of the Oman project, which involves an examination of the entire terrain of an almost trackless country, is a measure of how far landscape architects have come from their origins as glorified gardeners.

According to Platt, the job of the landscape architect is to join that technological capability with a sense of esthetics and respect for the environment to create "high-quality design and a good bottom line."

Landscape architects, he said, don't like to think about money, and the designers of buildings often don't like to think about vegetation, stream runoff, road access routes and enhancement of the natural environment.

"If the architect is only interested in buildings," he said, "you get a lot of junk. If the landscape architect is just thinking of vegetation, you get a lot of frilly posies. We don't do little picturesque gardens out of context. We like to be sensitive to the environment we're working in and see land used properly."

SWA, which employs more than 100 full-time landscape architects and has annual billings of about $7 million, is one of the nation's largest and most influential firms of its kind. From offices in California, Texas, Florida and here, its architects work with such well-known firms as Skidmore, Owings & Merrill and Eero Saarinen Associates on-site planning and landscaping of parks, urban centers, office parks, college campuses and ski resorts.

The firm's stamp is on such mammoth projects as the City of Irvine, Calif., the Golden Gate National Recreational Area and the industrial city of Jubail, Saudi Arabia, and on individual buildings as mundane as a brewery in Fairfield, Calif. One of its most ambitious current undertakings is the redesign of the bleak, freeway-scarred waterfront of San Diego.

Way's photogrammetric technique is useful in large-scale projects dealing with undeveloped land. Much of SWA's work, however, is with urban projects in relatively cramped spaces.

"We opened our Boston office because we're interested in urban projects. People in cities should have a respect for their environment," Platt said. But SWA was unsuccessful in two recent attempts to obtain important contracts in Washington. The firm was a losing bidder on the redesign of the area around the old Post Office on Pennsylvania Avenue NW, and was a member of one of the losing syndicates bidding for the contract to develop the Portal site near the 14th Street bridge.