Page images

Technologies: The
Role of Mainframe
Computing in the U.S.
Geological Survey
By Gayle Gordon, Tom
Faulds, and Joe Aquilino

Mainframe computers have played a significant role in the U.S. Geological Survey since the first computing equipment was installed in 1947. That role has shifted over the years as new and powerful alternative computers have become available. Despite that shifting focus and the prolifer ation of minicomputers and desktop microcomputers, the computing technology requirements study, conducted by the USGS in 1986–87, reaffirmed the bureau's continuing need for mainframe computers.

In 1956, the Survey procured its first mainframe computer to process scientific data. Over the years administrative applications, as well as scientific work, were transferred to computers, where the functions could be performed faster. Research that had previously been laborious or even unimaginable was now possible. Computers were able to process large quantities of data, rapidly perform complex calculations, and provide storage and retrieval of large volumes of data. These same functions can be and are performed on the Survey's current mainframe computer, but now they are done faster, and a significantly larger amount of data is processed and stored Computer technology has changed so rapidly that today many microcomputers located on desktops throughout the organization can perform better than those first Survey computers, which occupied many square feet of space.

Large mainframe computers, however, still offer advantages to USGS scientists and administrators. First, they have sophisticated software such as data base management systems that are able to take advantage of both the power of the main frame processor and its capacity to store and process large quantities of data. For example, the mainframes in the National

Center, Reston, provide storage and retrieval for two of the Survey's largest data bases, the Water Data Storage and Retrieval System and the National Digital Cartographic Data Base. Providing access to large data bases is a key role for the mainframe computers today and will probably continue to be so into the future. By having these data bases stored and maintained in a central location available through a communications network, scientists and administrators throughout the organization are able to access the data and either manipulate the data on the central computer or download it to their own minicomputers or microcomputers.

Mainframe computers provide highspeed processing. A model that takes hours to run on a minicomputer or a microcomputer, for example, can often be run in a fraction of the time on a mainframe. Also, many jobs can be run at the same time without one affecting the performance of the others. During prime-time working hours, for example, the Reston mainframes average more than 50 concurrent interactive sessions. The shift from the former single-job, batch processing to the current interactive computing is one of the most significant changes in the last ten years of mainframe computing, and its effect is expected to increase in the future.

Another glimpse into the future suggests that even higher capacity storage devices will be added to mainframe computers. In 1987, the Information Systems Division (ISD) installed a tape cartridge system on one of the mainframes. This technology is more reliable than magnetic tape and provides great savings in the amount of space required to store a comparable amount of data. It is anticipated that in the future more powerful storage devices such as optical disks will be added to ISD's mainframe configuration. This example of sophisticated peripheral devices illustrates another advantage of centralized computing. That is, advanced and costly technology such as storage devices, printers, and plotters can be made accessible to a large number of people through computer communications capabilities.

ISD's Data Center in Reston each month processes 35,000 jobs, mounts and processes 10,000 magnetic tapes, and prints and distributes 1 to 1.5 million pages of print. Because of a predicted 10-fold workload increase from 1987 to 1992, and because the existing systems are old and expensive to maintain, the bureau-wide Information Systems Council, with the concurrence of the Director and Executive Committee, has agreed to proceed in fiscal year 1988 with the acquisition of a new computer processing unit to replace the existing systems. The new system will significantly improve processing of today's workload and will help ISD support USGS computing needs in the future.

· Developing a system to provide detailed tracking, accounting, and reporting of foreign travel for the Office of International Geology. • Cooperating with other Department of the Interior bureaus on projects involving forestry management, environmental impact, and safety management accounting and reporting. • Providing continued assistance to the Department of the Interior Office of Surface Mining Reclamation and Enforcement and their contractors in the application and use of the Model 204 data base management system supporting Office of Surface Mining development of a large data base for lease applications, administration and control, and mining violations. • Experimenting in the use of standard statistical and graphics software for performing geographic information system manipulation and analysis of geographically referenced data, such as land use and land cover, and radon level measurements.

Applied Technology
By Tod Huffman

The Compact Disc Revolution By Rick MacDonald

An important aspect of the Information Systems Division's support services is providing systems analysis and developing applications software to the U.S. Geological Survey. During fiscal year 1987, ISD provided automated systems and technology development services to the full range of administrative and scientific, research, and technical programs of the Survey. These automated systems and development services included: • Developing and implementing a nationwide automated system to initiate, monitor progress, and complete action for all Survey requests for Personnel Actions, SF-52's. • Developing and implementing an automated system that permits the Survey's personnel office to compile a data base of nationwide vacancy announcement information, and to index, query, and automatically distribute those vacancy announcements nationwide. • Modelling the use of modern relational data base management systems for the storage, processing, and retrieval of National Mapping Division digital cartographic index data and Geologic Division geophysical/geochemical coal resource data. • Staffing and equipping the division's Graphics Laboratory, which provides the USGS with a wide selection of graphics hardware and software and offers support and consultation services on the use, design, and preparation of computerrelated graphics.

A remarkable new technology is about to revolutionize the way in which the U.S. Geological Survey disseminates its vast archives of earth-science data. That technology-CD ROM (compact disc read-only memory)-embodies some noteworthy characteristics with respect to data storage and data distribution. First, the disc itself is small, compact, and virtually indestructible. It is physically identical to the audio compact disc, which is currently revolutionizing the consumer audio marketplace. The disc itself can hold between 550 and 600 megabytes of data and can provide random access to any block of that data in less than a second. Figure 1 shows equivalent storage capacities of the CD ROM and of more traditional media. The disc drive that holds the CD ROM disc can be physically connected to a personal computer and costs between $500 and $600.

Data to be placed on the CD ROM can come from many sources. Data can be

[merged small][subsumed][subsumed][ocr errors][ocr errors][merged small][merged small]

prescribed format, then “mastered” onto a glass CD ROM disc (in much the same manner as an audio recording) at a cost of about $3000. Once the master is created, duplicate CD ROM discs can be “stamped out” at a cost ranging from $30 down to $5 each, depending on quantity. Both the mastering and duplicating are available at over half a dozen CD ROM mastering plants open in the United States. The overall process is depicted in figure 2.

Several successful prototype projects conducted by the USGS over the past year have proven that CD ROM is a costeffective means for distributing digital data. The first of these prototype projects was developed to assess the usefulness of CD ROM technology to a broad spectrum of users. The prototype CD ROM disc that the Survey produced contained samples of both administrative and scientific data bases, including the Cartographic Catalog, Earth Science Data Directory, Federal Information Resources Management Regulations, Federal Acquisition Regulations, geographic reference database, Geographic Names Information System, geologic index, Geological Long-Range Inclined Asdic (GLORIA), Geographic Information System (GIS) Files, Mineral Resource Data System, National Uranium Resource Evaluation, PUBMANUS (listing of published USGS reports and maps), Rock Analysis Sample System, Seismic Data, the USGS Manual, WRSIC (Water Resources Scientific Information Center abstracts), National Water Summary data, and streamflow gaging station data.

The USGS is currently investigating microcomputer-based search and retrieval software that lends itself to use with CD ROM. Related to this investigation, part of the first prototype CD ROM disc was devoted to testing a half dozen of the leading software products, all of which use the same pair of reference data bases provided by the General Services Administration. Both the data bases and the retrieval software packages were placed on the prototype discs, which were then distributed, along with the necessary readers and instructions, to a broad cross section of USGS scientists and other professionals. Testimonial information was collected on the usefulness and effectiveness of this new technology.

This initial experience and the testimonies from the users enabled the developers

[blocks in formation]

of the prototype disc to establish some basic guidelines. If a data base is relatively large and relatively static, has a potentially large audience of users, and lends itself to processing on a microcomputer, it may be an excellent candidate for CD ROM technology. With these criteria in mind, marine geologists from the Survey's deep sea-floor mapping project agreed to place their data from the GLORIA sidescan sonar system onto a CD ROM disc for data access and distribution.

An important consideration in placing any scientific data base onto CD ROM is the potential market that can be created for the data. When the GLORIA data base of digitally captured sea floor mosaics is available on CD ROM, the potential community of users will no longer be restricted to those individuals or organizations who have access to minicomputers or mainframe computers but instead would also be open to those who have desktop microcomputer workstations.

The CD ROM that contains the GLORIA data not only exhibits the ideal characteristics of a candidate for this technology but also exemplifies the opportunities for interagency cooperation afforded by such a large-capacity, inexpensive storage medium. The National Oceanic and Atmospheric Administration is a welcome partner in the project as they develop indexing software for the disc, and the National

Aeronautics and Space Administration has contributed image processing software to the effort.

As the USGS continued to develop its expertise in CD ROM, it became clear that other Federal agencies could benefit from the Survey's experience. Therefore, the USGS formed a special interest group called SIGCAT (Special Interest Group on CD ROM Applications and Technology) to pursue the many aspects of this technology and, in doing so, to establish a forum for the exchange of ideas, information, and experiences for the benefit of all concerned.

SIGCAT, which held its first official meeting in 1986, now has close to 1,000 members, representing over 120 different Government organizations. About 75 percent of the membership is from government and 25 percent from private industry. Several individual working groups have been formed under SIGCAT to focus on the issues of search software, drives and media, library and educational applications, and CD ROM projects ongoing in the Federal sector. Recently, the USGS together with the National Bureau of Standards established a CD ROM Lab, where the organizations involved in SIGCAT can evaluate the various CD ROM hardware components available in the marketplace. Such evaluation should prove useful to agencies that are in the process of investigating CD ROM technology.


Figure 3. An 8-bit RGB color space. A, Color cube for cube number 255. B, Color cube for cube number 127, which is theoretically inside cube number 255.

« PreviousContinue »