Page images

production system. Attaining this goal of modernizing the National Mapping Program will enhance the USGS's ability to meet national requirements for up-todate multipurpose cartographic data and map products.

The concerns of both scientists and policymakers are now focused on the potential for long-term


Geographic Information Contributes to Studies of Global Change

By Raymond D. Watts

titatively. The concerns of both scientists and policymakers are now focused on the potential for long-term changes that will span decades or centuries and involve the entire, tightly coupled, land-oceanatmosphere global environmental system.

The land component participates in global environmental processes in many ways. It is a new challenge to geographers and cartographers to describe and to measure the role that land plays in order to provide quantitative information that is useful in the development of computer models of the global environment. One of the major challenges to scientists who study global change is to identify the relative roles of human and natural processes and the ways that environmental change, once begun, feeds back through natural and human behavior to effect further change.

The USGS is undertaking studies of the land relationships indicated by red lines in figure 2. The central question of these studies is how to provide descriptions of land characteristics that will serve

The summer of 1988 was hot and dry. Farmers in the northern Great Plains endured massive crop failures, and Washington, D.C., sweltered in the hottest summer since the 1930's. Political attention was focused on concerns about the effects of human alteration of the global environment, including its climate.

In congressional testimony, one scientist expressed “99 percent confidence” that the weather of 1988 was attributable to enhanced greenhouse heating of the atmosphere-a result of higher atmospheric carbon dioxide concentrations from the burning of fossil fuels and deforestation. Most scientists, however, would not agree that such a direct connection has been established. In spite of great progress, current understanding of many aspects of climatic behavior remains rudimentary.

Today's computer models of climate are just beginning to incorporate realistic oceanic behavior, and their descriptions of land-atmosphere interactions are woefully inadequate. On short time scales (days to weeks), the atmosphere, oceans, and biosphere are relatively independent. On slightly longer time scales (months, seasons, years), the atmosphere, oceans, and biosphere affect each other, but feedbacks are relatively unimportant. On yet longer time scales (years, decades, centuries), all parts of the Earth's environmental system interact and affect each other-but in ways that scientists do not understand qualitatively, much less quan


Figure 2. Relationships of human activities and natural processes acting on the land, effects of land changes on natural processes and on humans, and feedbacks into further land change. The relationships represented by red lines are to be examined by the National Mapping Division for application to interdisciplinary global studies.


Climate and


dynamics Solid-earth processes


Effects on Humans -7

tions and others will provide an enduring resource for scientists who study the function and changes in the global environment.

The Digital Topographic Map of the Future

By John E. Findley

the needs of scientists who study global environmental change. Ecologists, meteorologists, atmospheric chemists, and scientists from numerous other fields need information about topography, geology, botanic cover (including seasonal variations), soils and soil development, erosion, and other land characteristics. Global modeling studies require information different from that needed for local process studies. For geographers, cartographers, and others who build essential data bases, an understanding of resolution and accuracy requirements is a prerequisite for an effective contribution to the interdisciplinary global research program.

Computer models of environmental processes are a distillation of current scientific knowledge. They are quantitative by nature and design; they describe, for example, reservoirs, flows, temperatures, and other characteristics numerically. Limitations in computer size and speed force modelers to characterize the land, its features and behavior, in aggregates. The complexity of our immediate neighborhood, which is so familiar, is therefore lost when a regional representation is formed. Few studies have been done on the quantitative validity of such generalization. The National Mapping Division is studying the problems of quantitative scaling of environmental observations and descriptions as part of its effort to develop strategies for observing and describing the land portion of the global environment. The strategy-development efforts are, themselves, interdisciplinary studies done in collaboration with scientists of many agencies, institutions, and disciplines. Once the strategies are developed, the data gathering will also be interdisciplinary, multi-institutional endeavors.

The USGS's EROS Data Center is home to a 20-year archive of Landsat images (both digital and film) as well as to millions of scenes of satellite and aerial photographs that provide a unique set of historical environmental observations. The Data Center's activities are expanding considerably in support of interdisciplinary studies of global change. The Data Center will become the long-term archive for data obtained by NASA's Earth Observing System, Eos-set for launch in the mid-1990's. These observa

Scientists in many different disciplines, both in government and the private sector, use and depend upon reliable earth science data. Until the mid-1970's, analog data in the form of printed line maps and image products were the sole source of base cartographic data. More recently, base cartographic and other earth science data have been made available in digital or computer-generated form. These data were initially developed for use only by scientists and were typically distributed on 9-track computer tape without supporting software. The user community, therefore, was effectively limited to persons having some background in the earth sciences (particularly a working knowledge of coordinate geometry and transformations), a reasonable level of programming skill or access to programming resources, and access to a mainframe or large minicomputer.

Before microcomputers and the development of the geographic information systems industry, these limitations were not significant impediments. Few users possessed the knowledge to effectively use the data, and even fewer had access to adequate computing resources. Today, however, with the phenomenal computing power that is readily available in the microcomputer market at reasonable prices, and the flexible data integration and manipulation software that is available from numerous vendors, the potential user community of earth science data has become enormous. The USGS and other suppliers of digital spatial data need to evolve with the user community and with applicable technology and, accordingly, to provide data in


forms compatible with emerging hardware and software systems.

Toward this end, the USGS is investigating several novel approaches for providing users with reliable and accurate data on a number of different media. One such approach is an ongoing research effort called the “Digital Topographic Map of the Future,” the objective of which is to develop prototype models of advanced digital data sets on CD-ROM (Compact Disk-Read Only Memory) that depict county-based geographic information. This project proposes to test not only a novel medium for data distribution (CD-ROM) but also a novel spatial coverage (county versus the traditional quadrangle base) and a novel marketing approach in which data display and processing software will be packaged with the data. The goal is to develop a distribution package that will permit novice users to create or display custom maps from standard National Mapping Program digital data by using personal computers. The initial test CD-ROM will contain software to allow tally rectified by using a digital elevation users to display digital line graph, digital model to remove displacement of feaelevation model, Geographic Names

tures in the photograph due to changes Information System data, and orthopho in elevation. The resulting digital orthotoimages of Arlington County, Va.

photoquad accurately portrays ground
distance and the spatial relationships
among objects in the image. A major
advantage of the digital orthophotoquad
is that advanced computer processing,
such as image enhancement, data merg-
ing, and digital mosaicking, can be
applied to improve the interpretability of
the photograph, to allow other geo-
graphic and cartographic data to be
simultaneously considered, and to permit
adjacent photographs to be joined into a

single larger map.
The USGS is the Federal agency that In a pilot project with the Soil Con-
is responsible for providing much of the servation Service, Department of Agricul-
spatial information that is used by scien ture, the USGS prepared 140 3.75- by
tists to review, analyze, and evaluate geo 3.75-minute digital orthophotoquads
graphic information. In the past, this (quarter-quadrangles) at 1:12,000 scale
spatial information was conveyed primar over Dane County, Wis. Both the scale
ily by aerial photographs, line maps, and the location were selected as being
data digitized from maps, and standard the most useful for a number of potential
orthophotoquads. The USGS has devel applications. Samples of corresponding
oped advanced techniques to produce a soft-copy products also were produced
new product termed a “digital” orthophoto- for this project for evaluation by the Soil
quad from aerial photographs.

Conservation Service. Several additional To produce these digital orthophoto prototype products are being prepared to quads, computer-generated files, created determine the usefulness of orthophoto by scanning aerial photographs, are digi quarter-quadrangles for selected areas,

Digital orthophotograph of Black Earth, Wis., created by using advanced computer techniques.

Large-Scale Image
By John E. Findley

[ocr errors]

especially those urban and suburban areas that are experiencing rapid change and development. The USGS and the Soil Conservation Service also are evaluating the use of digital orthophotographs in geographic information system applications. Continued development of both hard-copy and digital orthophoto products will provide a valuable addition to the USGS’s National Mapping Program, particularly the National Digital Cartographic Data Base.

to use in preparing the extensive number of maps for the 1990 decennial census.

Work on the pilot project began in June 1989. The USGS provided land use information for the Agricultural Atlas by using the data from its 1:7,500,000-scale “Land Use" map in the National Atlas of the United States of America. When the pilot project began, no digital data were available for this map; only the map separates used in the original printing existed. In the mapmaking process, information to be portrayed on the map is color separated into individual sheets that are used in the offset printing process. To capture the graphic data from the original map separates in

U.S. Geological
Survey-Census Bureau
Continues — Production
of Agricultural Atlas

From this pilot project the two agencies will

determine the most effective method to use in preparing the extensive number of maps for the 1990 decennial census.

By Loreen G. Utz

When the Nation's census takers have finished gathering the information on the population of the United States, the work on producing the results of the 1990 decennial census will have just begun. Maps will be needed to help statisticians, demographers, policymakers, and others use the myriad of information from the census.

The USGS expects to plot data for more than 6,000 maps during the 3-year period following the 1990 census in continued support to the Bureau of the Census.

The USGS has been working with the Census Bureau since 1983 in preparation for the 1990 census. As part of an amendment to that working agreement, the two agencies began a cooperative pilot project in fiscal year 1989 to test methods for producing publicationquality maps from digital data. By using information from the recent Census of Agriculture and land-use information and film plots supplied by the USGS, the Census Bureau will produce the 1987 Agricultural Atlas of the United States. From this pilot project the two agencies will determine the most effective method

digital form, the USGS used a method called raster scanning, in which the lines and other markings on a map are converted into computer-readable form. Photographic methods were used to combine the separate map sheets to make a color proof for each of the 16 different land use categories. Each of the color proofs was scanned and edited to remove extraneous detail. The resulting raster data files were sent to the Census Bureau, where they were subsequently converted to a format compatible with Census data.

Over 300 maps-about 200 dot distribution maps and about 100 choropleth (thematic) maps — will be produced for the 1987 Agricultural Atlas of the United States. Dot distribution maps consist of dot symbols representing the collected statistics. Because the data were collected by county, each county will be portrayed with an appropriate number of dots representing the density of the data. Evenly distributed dots throughout each county would mislead map users into believing that the data are also evenly distributed. To present a more accurate portrayal, the Bureau of the Census will apply landuse information to influence the distribution of the dots. For example, if a county produces 435,000 bushels of wheat and 1,000 bushels are represented by 1 dot, then 435 dots would be placed in that county. However, any land within that county that is classified as urban or woodland would receive no dots because one would assume that no wheat would be grown in those areas.

The choropleth maps will consist of county units, colored to create lighter or darker areas in direct proportion to the density of distribution of the theme subject. Each county is treated as a single unit without regard for the actual location of the theme within the county.

The production of the 1987 Agricultural Atlas of the United States will serve as a prototype for the production of the thousands of maps that will result from the 1990 decennial census.

Side-Looking Airborne
Radar Glacier
By John E. Jones, Bruce F.
Molnia, Robert M. Krimmel,
and James W. Schoonmaker, Jr.

ter of the glacier and up to 3.0 miles a year in localized areas of the glacier during surges. A team of scientists visited the glacier in September 1988 to investigate both the causes of the discrepancy between ice-flow velocities and stationary patterns and the characteristics of the glacier's surface corresponding to the backscatter patterns.

Investigators found that the patterns on the radar images generally correspond to adjacent topographic highs and lows of the glacier's surface. The bright signatures are usually topographic highs that are characterized by extensive crevassing; the dark signatures are topographic lows having few crevasses. Because the patterns remain stationary while the ice moves across the patterns, the investigators hypothesize that the glacier's surface features mimic the configuration of the glacially and fluvially eroded bed at depths as great as 3,000 feet below the ice surface. Like standing waves in a flowing stream, the topographic lows, or swales, appear to be areas of compressional flow, while the topographic highs are characterized by extentional flow.

Airborne and satellite radar detect the features because crevasses and large hummocks on topographic highs act as radar reflectors and backscatter a large portion of the signal, whereas water and snow in the smoother surfaced topographic lows reduce the return signal. Scientists plan to conduct ice-penetrating radar surveys to develop this concept further. The investigation is also being expanded to Bering, Hubbard, and Grand Plateau Glaciers, all located along the coast of southern Alaska.

These observations and hypotheses have important implications. Because crevassed and uncrevassed zones can be identified by using radar data, safe landing sites and transportation routes can be mapped to facilitate geologic exploration. This technique was used successfully during the 1988 field investigation of Malaspina Glacier. Further, if the surface topography on this glacier and other ice sheets is an expression of subglacial morphology, then it may be possible to develop a fluid dynamics model for estimating ice volume in support of globalchange investigations. This model would use remotely sensed data to measure ice-sheet area, flow velocities, wave

A USGS study of Malaspina Glacier found that certain bands of airborne and satellite radar images show large patterns of complex bright and dark radar backscatter on the surface of the glacier. These patterns, 0.3 to 6.0 miles in length, create a visual image that resembles bedrock features and the topography of nearby mountains, such as cirques and drainage networks. When inspecting airborne radar images from 1976, 1980, and 1986 and Seasat satellite radar images from 1978, along with topographic maps prepared from aerial photographs flown in 1958 and 1978 and other data sets, scientists found that the patterns have been stationary for nearly 30 years. This finding is unusual because ice-flow velocities of Malaspina Glacier are typically 0.1 mile a year near the cen

« PreviousContinue »