Environmental Sciences: Submitted by Robert Davis
This document provides brief responses to the following questions
posed by the UVA Faculty Senate:
1. What do you see as the clearest gains and losses brought thus
far by the use of networked computers at UVa?
2. What do you see as the greatest possibilities, limitations, and
dangers these machines pose in the future?
3. What priority should the university place on information
technology?
4. Please add any additional comments you would like to call to
our attention.
These broad questions deserve careful consideration. What follows
are several relevant points being discussed within the Department of
Environmental Sciences.
1. Benefits of Network Computing
Data Access. Among the most significant benefits of network
computing to University of Virginia students, staff, and faculty is
ready access to data and information. Local-area networks within the
University, wide-area networks, and the Internet all contribute to a
revolution in the availability of data. Data and information that,
just a decade ago, were the province of specialized agencies or
laboratories are now publicly available with near immediate access.
For example, in the environmental sciences, the limited distribution
of satellite remote sensing data once restricted their use to a few
centers of the National Aeronautics and Space Administration and
major university programs. Within the next year, imagery collected by
NASA's Earth Observing System will be available to all on the
Internet along with numerous derived products that will provide
insight without specialized training. At the University of Virginia,
students and faculty now access near real-time weather observations
and forecasts, and the entire suite of observations relevant to
global environmental change can be accessed on the Internet.
University of Virginia scientists conducting investigations at the
Virginia Coast Reserve Long-Term Ecological Research Site make their
data available to broad communities on the Internet. Access to data
no longer restricts research interests in the environmental
sciences.
Communications. Electronic mail and related network communications
are enabling interaction and collaboration that were impractical not
long ago. University of Virginia faculty are able to interact with
colleagues worldwide. It is now common to co-author papers through
network exchange of drafts numerous times within a single workday.
Visualizations of results, even with animations, can be available to
students and colleagues as they are derived, whether from
instruments, field or earth observations, or by evaluating
mathematical models. Computer network communications are enabling a
variety of virtual institutions both within the University of
Virginia and across continents.
Distribution of Teaching Materials. The World Wide Web interface
to networks has replaced photocopy as the primary means of
distributing materials to students. The additional facility for
interaction between students and with instructors reaches far beyond
the Kinko's class notes of a decade ago.
University Administration. By design, the University comprises
diverse departments, schools, and colleges spanning near the breadth
of human activity. Effective linkage of these has challenged
University leaders and administrators for decades. The complexities
of student life have distracted numerous young scholars, and
virtually all faculty and staff complain about the burdens of
University administration. The distribution of information, including
catalogs, course schedules, calendars, and policies on computer
networks is the beginning of improvements in University
administration that will be enabled by networks.
2. Network Computing Concerns
The distribution and processing of data and information on
networked computers is expanding so rapidly as to distract from
traditional methods of research and analysis. Increasingly, students
turn to the World Wide Web as a primary information source. This
interface can appear more attractive than reading books and journals
or plowing through government documents and records. It is not
uncommon for students to report having exhausted available
information sources without consulting a single printed document.
Ironically, the allure of the World Wide Web and the Internet
distract attention from the fundamental computer sciences including
scientific and engineering computation and visualization. While
various University organizations rush to provide training in the
access and distribution of information through the World Wide Web
interface, short courses in programming and the operating systems
that enable scientific and engineering computation are disappearing.
Increasingly, students are able to construct dramatic Web pages but
cannot program even basic procedures for processing data and
information. Although formal course work is critical, mentoring has
always been a key aspect of training in computer programming and
application. The current focus on the World Wide Web has created an
imbalance in the exposure of students to computing topics through
peers and less formal training. Networking enables access to
information and computational resources on a variety of technologies.
This versatility is particularly important to meeting the needs of
the diverse University community. But network efficiency and
simplified system administration argue for hardware and software
standards including personal computers, workstations, operating
systems, and application software. Such trends toward a common
denominator can seriously erode the quality of computational
resources available to the many disciplines of the University.
Furthermore, any move toward standardization of desktop computing
hardware or software runs counter to current national trends and
developing technologies that allow for information retrieval that is
completely independent of the computational platforms being used.
The focus on distributed computing at UVA also distracts attention
from the critical contributions of large high-performance
computational facilities. In principle, computer networks can provide
access to high-performance computers independent of their locations.
But the development and operation of distributed systems is
demanding, and the maintenance of links to super-computers is easily
assigned low priority by systems managers and engineers chosen for
their interest and skill in distributed computing and information
technology. Capabilities for high-performance computing are deficient
at the University of Virginia. This seriously impacts the ability of
science and engineering students to engage in numerous aspects of
research.
3. Information Technology Priority
The University must rank information technologies among its top
priorities.
4. Additional Comments-Personal Computers and Workstations
The increasing significance of computer networks within the
University of Virginia is coupled to several characteristics of the
modern computers they link. The most important of these is the power
and performance of affordable and small personal computers and
workstations. Virtually every student, faculty, and staff member has
access to a personal computer with substantial computational power
and versatility. It is not uncommon to find several computers in the
offices and laboratories of faculty, and students use a variety of
computers in University facilities. Numerous students and faculty
carry small laptop computers wherever they go with the computational
power of the massive mainframe systems of a decade ago.
In summary, network computing has literally revolutionized the way
in which we do our science. It has vastly broadened our access to
information, improved our ability to interact with colleagues
globally, and provided us easy access to the data we need to do our
research. Science is discovery, and computing has become instrumental
in that discovery process.