Citation
Geographic information systems : tools of emerging importance for landscape architects

Material Information

Title:
Geographic information systems : tools of emerging importance for landscape architects
Creator:
McGovern, Ruth
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Master's ( Master of landscape architecture)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
College of Architecture and Planning, CU Denver
Degree Disciplines:
Landscape architecture
Committee Chair:
Young, Daniel B.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
Copyright Ruth McGovern. Permission granted to University of Colorado Denver to digitize and display this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.

Downloads

This item has the following downloads:


Full Text
GEOGRAPHIC INFORMATION SYSTEMS: TOOLS OF EMERGING IMPORTANCE FOR LANDSCAPE ARCHITECTS
I
AN INTRODUCTION TO THE TECHNOLOGY AND CRITERIA FOR EVALUATING AND IMPLEMENTING A GIS AT UCD COLLEGE OF DESIGN AND PLANNING
BY
RUTH MCGOVERN
A view of Mt. McKinley from the Northeast produced by the Comarc system.


GEOGRAPHIC INFORMATION SYSTEMS: TOOLS OF EMERGING IMPORTANCE FOR LANDSCAPE ARCHITECTS An Introduction to the Technology and Criteria for Evaluating and Implementing a GIS at UCD College of Design and Planning
by
Ruth McGovern
A thesis
submitted in partial fulfillment of the requirements for the degree of Masters in Landscape Architecture The University of Colorado at Denver May 1984


THIS THESIS IS SUBMITTED AS PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR A MASTER OF LANDSCAPE ARCHITECTURE DEGREE AT THE
UNIVERSITY OF COLORADO AT DENVER COLLEGE OF DESIGN AND PLANNING GRADUATE PROGRAM OF LANDSCAPE ARCHITECTURE
(Committee Member's Name & Title)
___Odober 19. 1964
DATE ^


CONTENTS
LIST OF FIGURES.............................................. i
I. INTRODUCTION ...................................... 1
A. Statement of the Problem........................... 5
B. Purpose of the Project............................. 7
C. Procedure for the Investigation.................... 7
II. OVERVIEW OF GISs................................... 11
A. GIS Components.................................... 12
B. Overview of Representative Projects............... 14
1. A Hiking Trail Suitability Model ......... 14
2. The Northwest Colorado Council of
Governments Water Quality Management
Plan....................................... 18
C. Information Flow.................................. 28
1. Data Gathering............................... 29
2. Data Entry................................... 31
3. Data Structure............................... 34
4. Editing and Updating......................... 37
5. Storage of Data on Tape or Disk........... 39
6. Data Retrieval, Analysis, and
Manipulation .............................. 40
7. Output Formats and Output Devices......... 4 3
D. Classifications of Systems ................... 46
1. Orientation of the System.................... 47
2. Size of the System........................... 49
E. Considerations for LAs............................ 50
1. Advantages
50


52
54
55
56
57
59
59
61
62
63
64
65
67
68
68
70
70
71
74
75
78
82
82
Disadvantages Scale. . . .
2.
3 .
4. Accuracy .................................
5. Cost .....................................
6. Rethinking Our Old Methods ...............
CRITERIA FOR EVALUATING GISs .....................
A. Establish Your Needs and Determine
Your Requirements...........................
B. Determine Appropriate Size and Orientation
of the System...............................
1. Size of the System .......................
2. Orientation of the System.................
C. Evaluate Reliability of the Vendor ...........
D. Appraise the GIS for Ease of Use and
Flexibility.................................
E. Cost Considerations...........................
RECOMMENDATIONS TO UCD LA PROGRAM.................
A. Needs/Requirements of the UCD LA Program . . .
B. Appropriate System Size and Orientation
for UCD.....................................
1. Orientation of the System.................
2. Size of the System .......................
3. Cooperation with Public and Private
Agencies ...............................
4. Incorporation Into the Curriculum.........
SUMMARY AND CONCLUSIONS...........................
APPENDICES .......................................
A. Survey of the Profession ......................


B. Glossary of Terms............................. 84
FOOTNOTES.................................................... 9 0
BIBLIOGRAPHY ................................................ 93


FIGURES
Figure Page
1 GIS Applications 3
2 Process 8
3 GIS Hardware Components 13
4 Overlay 15
5 Hiking Trail Suitability Model 17
6 NWCCOG Region 12 19
7 NWCCOG Project Structure 21
8 NWCCOG Study Scale Conversion 19
9 NWCCOG Study Overlays to Define Landscape Units 25
10 NWCCOG Study Flow of Mapped Data 25
11 Information Flow 30
12 Data Relationships 32
13 Data Entry Techniques 36
14 Types of Spatial Data 36
15 Gridded Data 36
16 Slivering 38
1 7 Interactive Editing Features 38 >
18 Measurement Functions 41
19 Digital Terrain Analysis Functions 41
20 GIS Output 44
21 Criteria for Evaluating a GIS 60
22 Survey Results 83
1


CHAPTER I
INTRODUCTION
As landscape architects, we view our role in society as one of great responsibility. We charge ourselves with "stewardship of the land" (Fein, 1972), "design and management of the landscape (Killpack, 1980 ), and "dedication to the land and its people" (Marshall, 1973). This responsibility requires a thorough understanding of our biophysical and socio-cu 1 tura1 systems, as well as an understanding of the relationships and interactions of these systems. Getting an overall view of these relationships is no small or easy task, but is nevertheless one of the most important steps in our design and analysis process. The initial process of gathering, organizing, analyzing, and comparing natural and cultural information has been estimated to take as much as two-thirds of the total time involved in a project (Killpack, 1980). As we have more and more information to handle, it becomes increasingly important to improve our accuracy and efficiency. If we are to allow more time for thinking and creating, we must look to new methods of doing things .
While the increased complexity of our world creates problems, new opportunities are arising, in the words of Julius Gy. Fabos, to "restructure the form and mode of LA practice" (1983). Fabos suggests that landscape architecture is one of the


few remaining professions to escape the computer revolution. "Perhaps over ninety percent of our routine planning, design and management activities are still done in the way in which they have been done for decades.", he says.^ The computer is a new tool which has the potential to greatly aid and expand the field of landscape architecture.
One computer resource which has the capacity to be of tremendous help to the landscape architect is the geographic information system (GIS). A GIS is a computer system which is designed to store, process, analyze, and generate graphical output of information that is tied to a particular geographic location. In a GIS, locational data such as latitude and longitude are manipulated along with attribute data such as vegetation and soil types. Some special features for landscape architects are map overlays and composites, calculations of special areas and distances, and creations of buffer zones around specified areas. With the aid of a GIS, for example, an LA can build and display map models for site selection, or models for projection of development and change in an area. In addition, a GIS can perform such functions as delineations of view sheds, calculations of water flow and run-off, and calculations of density. Some other uses of GISs are listed in Figure 1. A GIS, then, can greatly reduce the time and drudgery involved in compiling this kind of data, increase the level of detail, generate outputs in forms which are attractive and easy to comprehend, and allow the landscape architect to focus the majority of his/her efforts toward design considerations.
2


SITE PLANNING
URBAN AND REGIONAL LAND USE PLANNING
LAND USE INVENTORIES
MASTER PLANNING
SOCIOECONOMIC STUDIES NATURAL RESOURCE PLANNING REGIONAL ENERGY IMPACT AND ANALYSIS
LOCAL EIS ANALYSIS
PREDICTIVE MAPPING
- PROJECTING POPULATIONS
- ESTIMATING RECREATION ACTIVITY
- PREDICTING STORM RUN OFF
- PREDICTING EROSION HAZARDS
ROUTE SELECTION FOR HIGHWAYS
ROUTE SELECTION FOR BUSES AND OTHER PUBLIC TRANSPORTATION
TOPOGRAPHY ANALYSIS FOR DEVELOPMENT
STATISTICAL MAPPING (CENSUS, CRIMES, ACCIDENTS, ETC.)
UTILITY FACILITY MAPPING AND MANAGEMENT
GIS APPLICATIONS
Figure 1
3


In spite of their apparently enormous potential for LAs, geographic information systems have been slow to enter the field of landscape architecture. At the recent Harvard Conference on Computer Graphics in Planning and Design, Chairperson Madson noted, "Geographic information systems have been widely developed and used in the fields of geography and cartography for more than a decade, but have been used very little in the fields of architecture and landscape architecture. The time and technology, however, are now ripe for the latter two fields."
Until recently GISs have been out of reach for all but institutions and agencies with access to large computers and the funding to afford the necessary large-scale data manipulations. Advances in both computer hardware and software have turned this situation around. Today, many useful GIS applications can be performed relatively cheaply using mid-sized (mini) and even small, desk-top (micro) computers. In addition, the conventional, time consuming and costly methods of formatting and inputting raw data into the computer (e.g., digitizing) are gradually being replaced by automated methods of data input such as scanning. Finally, GISs are becoming increasingly user-friendly and thus understandable and usable even for the computer novice.
Geographic information systems, then, are emerging as viable tools for landscape architects -- they have the capacity to provide an abundance of varied natural and cultural data from which we, as responsible managers and designers of the landscape,
4


should draw in making our design decisions. By taking much of the drudgery out of the process of assimilating this kind of information, as well as by significantly improving accuracy and efficiency, GISs can provide LA's with more time for thinking and creating; for generating and evaluating design alternatives.
A. Problem Statement
Although computer technology has advanced to the point where geographic information systems are affordable and have the potential to be tremendously valuable tools for landscape architects, GIS technology still remains largely inacccessible to members of the profession. There appear to be two primary reasons for this:
1) Virtually no information is currently available in the GIS literature which is specifically directed toward LAs. Learning about GISs and precisely what they can do for LAs often involves struggling through much jargon-riddled technical writing.
2) There are few outlets currently available to LAs for learning about and gaining hands-on experience with GISs .
The second problem of allowing LAs direct exposure to GISs is now being addressed at the university level, but so far only at a handful of universities on the east and west coasts. Here in Colorado, the need for LAs to make use of GIS technology is particularly strong given the problems we face with the rapid population growth along the front range. My survey of Colorado
5


professionals (see Appendix A), found that the current level of knowledge about GISs is quite low, but that interest in learning about GISs is high. Another significant finding was that virtually all LA firms either own a micro-computer or have plans to purchase one in the near future. Given these facts, it is apparent that the use of GIS technology by LAs is likely to surge in the near future if given the proper impetus. The LA program at the University of Colorado at Denver is in the unique position to spearhead this imminent surge because 1) it is the only accredited MLA program in Colorado, and 2) the Denver metropolitan area contains the state's greatest concentration of LAs.
By procuring all the necessary components of a user-friendly and versatile GIS, the UCD LA program would be able to offer students and area professionals a valuable outlet for learning how to make use of this rapidly emerging technology. This would likely have several beneficial effects:
1) It could spark increased utilization of GIS technology by LAs in the area.
2) It would raise the general level of computer literacy
%
for MLA students -- an issue of some importance to the degree program accreditation committee.
3) It would promote better linkage with those students and professionals in fields with whom LAs are associated who are making increasing use of computer technology.
4) It would anticipate the inevitable burgeoning of
6


5)
interest in GISs among LAs as the tools become more widespread. The system would already be in place in preparation for meeting increasing demands for GIS training.
It could serve as a resource for area professionals in much the same manner as a research library.
6) It would be a general asset to the university and could enhance the LA division's image.
B. Purpose of the Project
It is apparent that the time is now ripe for implementing at UCD just such a resource as I have suggested here: an easy-to-use and versatile geographic information system. The purpose of this paper then, is to help pave the way for this implementation by providing 1) an overview of GISs, written in terms understandable to those having no prior computer or cartographic experience, emphasizing the characteristics and practical applications of GISs as they relate to the landscape architecture profession, 2) a set of criteria and a format for evaluating various GISs based on specific needs, and 3) recommendations for implementing GIS technology at the UCD College of Design and Planning.
C. Procedure for the Investigation
The organization of this project is diagrammed in Figure 2. In presenting an overview of GISs, I will draw from several different sources -- from GIS literature geared toward computer specialists, geographers, and cartographers; from interviews with
7


PROJECT REVIEW
T==
INFORMATION FLOW
CLASSIFICATIONS OF SYSTEMS
OVERVIEW
OF
GEOGRAPHIC INFORMATION SYSTEMS

CONSIDER A1 LANDSCAPE / ‘IONS FOR ARCHITECTS
mmmmamsmmssB
CRITERIA FOR EVALUATING GIS
RECOMMENDATIONS
TO
UCD
LANDSCAPE ARCHITECTURE DEPARTMENT
PROCESS
Figure 2
8


various professionals knowledgeable about the systems, including GIS technicians, users, and vendors; and, finally, from my own personal experience in working with a GIS at WELUT (Western Energy and Land Use Team, a division of the U.S. Fish and Wildlife Service in Ft. Collins, Colorado). The overview section is structured from the general to the specific. The first portion of the overview illustrates the relationship of GIS technology to another computer graphic technology and defines the fundamental components of a GIS. The next portion reviews GIS projects
discusses the projects' formats and purposes, and explores the types of analyses used. The third portion of the overview goes more specifically into how a GIS works, exploring its functions and capabilities. This is designed both to provide a self-
to serve as a foundation for those individuals at UCD who may wish to make use of the ensuing section on GIS evaluation criteria.
The set of criteria in the third chapter of this report is intended to provide a framework for evaluating the suitability of GISs for LA applications. This is intended to provide a process for interested LAs to follow in evaluating and selecting a GIS. This framework is then utilized in this report as the basis for establishing goals and selecting a GIS for the LA program. The focus is, of course, on the LA Department, but as much as possible the goals and directions of other departments in the
of projects used by landscape architects,
contained introduction to GIS technology for interested LAs and
9


college are considered as well. Given the complexity of the decision-making process in determining an appropriate system, specific hardware and software recommendations are beyond the scope of this paper. Rather, the intention is to explicate the essential considerations and to explore various options available when choosing a GIS. It is hoped that this will provide the UCD Landscape Architecture Department with direction for implementing geographic information systems technology into the program, some initial impetus, guidelines for evaluating systems, and recommendations of issues for further consideration.
10


CHAPTER II
OVERVIEW OF GISs
To show how geographic information systems can be most useful to landscape architects, this section looks at how they work, how they are used, and examines some special features and problems related to the LA profession. Section A presents the components of a GIS. Section B looks at some actual GIS projects; examines their purposes, how they were structured, and how they were used. Section C describes in detail the flow of information in a GIS, including data input, storage, retrieval, analysis, and output. Section D discusses the requirements and special features of GISs as they relate to size and type of system. Finally, E sums up some considerations about GISs specific to the LA profession.
Many of the functions of a GIS overlap with those of a computer aided design (CAD) system. For example, both have data base management systems and extensive graphics modeling capabilities. The''difference between a CAD system and a GIS is that the focus of the CAD system is on design per se, whereas the focus of a GIS is on analysis and interpretation of geographically based data. The GIS is nevertheless integral to the design process in that it provides a context in which design considerations should take place.
GIS technology combines mapping and analysis capabilities so


that land use professionals can interact directly with geographic information and quickly obtain accurate answers to pertinent "what if" and "where" type questions.^ The questions might relate to such things as the best uses of a particular site, a comprehensive land use plan, or a coastal zone management program. GIS analysis may encompass mapped information, related tabular data, text information, and even satellite data.
A. GIS Components
A GIS may be a dedicated, stand-alone system, may be part of a network of computers used for GIS purposes, or may be only one of several functions performed on a computer. The principle hardware components of a GIS are shown in Figure 3. These include:
1) a computer,
2) data storage device such as a disk,
3) data transfer device such as a tape drive,
4) digitizer for encoding maps and aerial photographs,
5) graphics terminal (monochromatic or color), for data display and interactive communication,
6) line printer, used to print tabular data and generate low resolution graphics,
7) plotter for generating graphics "hard copies".
GIS software refers to the programs that perform the work requested by the user. The software contains 1) extensive data base management systems to handle data input, search, and
12


data atoraga lnd data tranafa
GIS HARDWARE COMPONENTS
Figure 3


retrieval, and 2) graphics capabilities to format data for output in the form of maps, charts, cross sections, and 3-D images.
B. Overview of Representative GIS Projects
This section reviews two diverse GIS projects, examines how the projects were designed and structured, and looks at some of the specific GIS functions that were used.
1. A Hiking Trail Suitabi1ity Model
The first project is a hypothetical one; it illustrates the use of a GIS to build a hiking trail suitability model. The final product from this project would be one map of an area delineating areas of high, medium, and low suitability based on a set of criteria established by the designer. It is used here because of its relative simplicity and because it illustrates a GIS function often used manually by LAs to build suitability models - the overlay.
The purpose of an overlay is to compare two or more maps or themes in a given area. For example, Figure 4 shows the result of overlaying a zoning map on a flood plain map. The GIS would create a new map showing exactly which portions of each zoning
•N
category were in each portion of each flood category. With the overlay function, maps can be produced which show 1) areas in common between two or more maps or data sets, 2) areas not in common between these maps or data sets, and 3) areas both in common and not in common on two or more data sets/ Such questions as how much national forest land lies in a given county
14


OVERLAY
Figure 4
15


or how much cleared land lies in a proposed development are simple examples that require an overlay manipulation.
Figure 5 diagrams the process that could be used to build a hiking trail suitability model. This is a standard process employed in setting up a strategy for building a suitability map from a set of overlays.
The first step is to define what the final map will be. The second step is to identify the components of the final map. In this case, erosion potential, hiking difficulty, aesthetic quality, and cost were identified as elements which would determine the ultimate suitability for a hiking trail. (Elements can be added or deleted depending on what the designer and client decide is important.)
The third step is to identify a set of sub-components. For example, views, noise, and surface texture were identified as sub-components which comprise the component aesthetic quality.
Fourth, additional sets of sub-components are again identified for each new category until one arrives at a data type or substantive map is. Determined as affecting views in this step, for example, were elevation, vegetation, and cultural features -each of whiph is available as a quantitative map.
To determine hiking trail suitability in an area, some subjective decisions need to be made. The first subjective decision is defining what components determine suitability. The second subjective decision is in ranking the four components said to determine suitability (i.e., which is the least/most important factor). To set up a ranking structure, weights or values are
16


HIKING TRAIL SUITABILITY MODEL
Figure 5


given to each of the components at each step. The final outcome, then, will be a map with areas weighted as to their hiking trail suitability and with each area having a quantitative value.
The model is executed in reverse order (i.e., right to left on the diagram), so that, in this case, elevation, vegetation, and cultural features are overlaid to make up the new map "views", which is overlaid with other maps to create the aesthetic quality map, which again is a sub-map of the final hiking trail suitability map.
2. The Northwest Colorado Council of Governments Water Quality Management Plan
The second GIS project reviewed here was an actual project involving a six county - 9,064 square mile - area in Colorado Region 12 (Eagle, Grand, Jackson, Pitkin, Routt, and Summit Counties). See Figure 6.
In 1976, Northwest Colorado Council of Governments (NWCCOG) contracted a GIS company, COMARC Designs Systems, to assist in preparing a Water Quality Management Plan for the six county region.
This GIS project is reviewed here because it 1) was a Colorado project, 2) is representative of the potential scale and scope of GIS projects, 3) is a completed project, with enough lapsed time to evaluate its overall effectiveness, 4) exemplifies the complexity and significance of initial project planning and structure, and 6) was subcontracted by landscape architect Ian McHarg. The project is first discussed in terms of the original
18


NWCCOG REGION 12
Figure 6
Grand County Land Usc Three Lakes Land Use
NWCCOG STUDY SCALE CONVERSION
______________Figure 8______________


goals, structure and final implementation. A summary reviews the project's strengths and weaknesses based on its original intent.
Cbjectives/Profile of the Region. The objective of the project was to devise a Water Management Plan to comply with Section 208 of the Clean Water Act. The plan was also intended to develop definitive, site specific water quality management policies for the entire area, to accommodate new legislation and guidelines, improved mitigation measures, and more refined data.^ The region's water resources are of national importance. It contains the headwaters of the Colorado River and headwaters feeding into the North Platte. Millions of people in several states and Mexico are dependent on these rivers for their water supply. Within the area are high mountain meadows, exposed bedrock summits, deep canyons, forested mountain slopes, valley meadows, cliffs, and broad intermontane basins. It encompasses scenic land, a national park, ski resorts, and deposits of valuable mineral resources. Demands for additional recreation, housing, and mining, coupled with a sensitive ecosystem, create a tremendous potential for conflict.
P lanning Process. Because of the size of the project, the amount of data, and the number of agencies involved, (the National Park Service,the National Forest Service, private landowners, and county planners), the overall project structure was complex. As shown in Figure 7, the main components were 1) inventory of data sources, 2) program definition, 3) data collection/encoding and formatting, 4) data analysis, 5) policy
20


NWCCOG PROJECT STRUCTURE
Figure 7
21


6
formulation, 6) comprehensive planning and 7) implementation.
iHXJ-Htor of data sources. The first step was a comprehensive inventory of data sources. An environmental atlas was prepared, documenting some 3000 sources of information on the region.
Program definition. Program definition was based on public input, policy requirements, review of potential activity areas, and design of seme prototype projects. It was recognized at the outset that water quality management must be viewed as a part of the overall planning effort. It was therefore necessary to consider other planning issues and the extent to which they would impact or be impacted by the Water Quality Management Plan. A comprehensive plan for the six county region, then, would be developed which would also look at development and growth for specific study areas within the region.
Data ££2lJl£££ i££/encod nef and f.££IB£££.ill2' From the inventory, and review of the data needed tc produce a regional water quality management plan, a regional data base was designed using new and existing data. U.S. Geologic Survey wss contracted,' for example, to perform a water quality inventory of surface water cf the region, while data cn soils information were gathered frem existing Soil Conservation Survey maps. All data were digitized and stored in the computer using C.onarc s Geographic Information System.:, COMPIS. Data encoded for the entire 9,064 square mile area included:^
22


for this was the overlay method described in the previous project review, the hiking trail suitability study. All the inventory maps were overlaid to determine all combinations of data and their frequency of occurrence. This was carried out in stages to allow for intermittent interpretation and analysis by the planning team. Figure 9 indicates this initial process of determining landscape units. First, the sub-basin maps were computer overlaid on the data base to break it down into nine separate sub-basins. Second, within each sub-basin, geologic formations were overlaid on the headwaters map to determine the occurrence of groups of geologic types in headwaters and nonheadwater areas; these were characterized as "Level One Landscape Unit". Third, these "Level One Landscape Units" were compared by the GIS to other maps in the data base to determine all the combinations of characterizations and extent to which each existed in the region. Finally, "Level Two Landscape Units" were defined. These represent the ultimate division of the area into ecological units -defining areas distinct from one another in
their hydrologic response to change.
\
Policy formulation. Figure 10 indicates the flow of mapped data through the computer. Each of the landscape units was evaluated in light of the regional water quality policies and was used as the basis for how policies should be applied to specific development sites.
Comprehensive planning. Public and advisory board meetings


Topography Geologic Formations Climate
Land Use/Cover Type
Potential Activities Zone (PAZ)
Soil Associations Surface Water Headwater Areas Sub-Basins
Reservoir Catchment Areas
Source maps had a variety of scales ranging from 1:1,000 to 1:250,000. Once the maps were in the system, coordinate conversions and scale changes were made by the GIS as necessary. Maps were merged with other maps and displayed with all the data at a consistent scale. From topographic data, the system computed slope, aspect, and elevation.
In addition to the regional data base, more detailed data bases were established for four specific study areas within the region: The Aspen-Snowmass Area in Pitkin County, the Lower Blue River Area in Summit County, the Three Lakes Area in Grand County, and the Walden-Coalmont Area of Jackson County. It was thought that these four areas faced most development,pressures and that they contained specific problems represented throughout the region. Therefore, their data bases were more comprehensive than the regional data base. A particularly valuable application of the computerized data base was extracting information for any locale by zooming into a specified area. For example, land use was encode'd for the whole region, but could be retrieved and displayed at a smaller scale. Figure 8 on page 19 shows land use being displayed for one county and then for one USGS 7-1/2' quadrangle within the county.
Data analysis. The initial analysis involved determining what ecological units occur in the region. The process employed
23


NWCCOG STUDY OVERLAYS TO DEFINE LANDSCAPE UNITS
_________________________Figure 9
NWCCOG STUDY FLOW OF MAPPED DATA
Figure 10


were held at nearly every stage in this project's process. In preparing the Comprehensive Plan in particular, the public became involved. A series of workshops were conducted in each county where the public was given the opportunity to comment on the proposed policies and then view those policies within the overall context of planning for their county. During these workshops, participants established criteria to be used for producing computerized development suitability maps. Using the Geographic Information System, they established population projections relating to specific growth inducements and locations. They developed phased growth projections for their county on the computer, resulting in a map of future growth in the region. Additionally, these maps were overlaid by the GIS onto the Level Two Landscape Units to determine potential conflicts between these growth patterns and each of the proposed water quality management policies. A report was then prepared for each county documenting these conflicts and listing mitigation measures which could be used.
Implementation. The original idea in utilizing a GIS for this project was to develop a regional comprehensive plan with several site specific development guidelines and to maintain a regional data base which could be used in subsequent years for other regional planning issues. It was thought that the regional data base could be used by different agencies, edited and changed through the years, and result in more consistency between various planning efforts.
26


While some of these goals have been partially or fully realized, a lack of foresight at key points initially caused problems in making use of the more site specific data. Housing development is in great demand in the area and was to be studied using the GIS data. A large percentage of the region is government owned (primarily U.S. Forest Service) land, so there are limited areas where development can occur. Another major limitation involves the installation of septic tanks. Soil constraints in certain areas prohibit development because of severe leaching. Although soil data were a part of the data base, they were on too gross a scale (standard Soil Conservation Survey maps) to apply to the five-acre lot size being reviewed in several county planning processes. In addition, soil profile and depth information was not included, rendering the data totally useless to the specific analysis needed for the sites' development review standards. Thus although the system could zoom into smaller areas for purposes of site specific analyses,
O
the base data was too coarse to allow credible use.
In the seven years since the comprehensive plan was devised, it has been used successfully on a county-by-county basis. Several county planners involved in the original study noted that the GIS was particularly useful for public meetings - both in promoting discussions via computerized future projection scenarios, and in credibility and clear documentation of the analysis process.
The original data base has been utilized and appended by outside agencies. A region-wide traffic study used and added to
27


various aspects of the data base. In addition, the data base was used by a utility company to determine the most suitable routes for transmission lines in the region.
Summary. This project exemplifies the potential scale and scope of GIS projects. The project used the GIS for large scale analyses, for promoting public participation, and inter-agency cooperation. It illustrates the use of various manipulative techniques typical in GIS analysis. Most notably, those include various topographic terrain analysis (slope, aspect, elevation), overlay and composite analysis, future projection analysis employing the "what if" query, map scale conversions, and locale "zooming" for analysis at smaller scales.
The project also demonstrates serious problems that can arise from a lack of forethought, underscoring the impprtance of initial project planning, of thinking through what the data will be used for and determining the level of depth and detail that will be required.
C. Information Flow
There 'are five distinct steps in the flow of information when using a GIS. These are: 1) data gathering, 2) entry of raw data into the computer (e.g., digitizing), 3) storage of digital data onto magnetic tape or disk, 4) selective data retrieval and analysis, and 5) output of processed data. Editing of data can take place between steps 2 and 3 and between steps 3 and 4. Often much of the pertinent data for a project will be available
28


already in digital form, in which case the time consuming efforts involved in step 2 can be reduced or eliminated.
Figure 11 shows these basic steps in the GIS flow of information. Each of the steps is examined in detail below.
1. Data Gathering
Data for geographic information systems can be gathered from a variety of sources: maps, aerial photographs, field surveys, assessor's records, census tapes, as well as from any standard source used by LAs in their analysis process. Since the data from standard sources need to be converted into machine readable (digital) form, (see Data Entry, below) the trend is toward using data already available in digital form. In the U.S., the United States Geological Survey (USGS) now distributes topographic and land use information in digital form, the Soil Conservation Service digitizes their maps, and LANDSAT (satellite derived data) is available on digital tapes.0 Many state agencies have converted their natural resource information to digital form as well. The amount and type of data used in a GIS, of course, is dependent on the scale and scope of the project and is determined by the users.
The spatial data of a GIS is most commonly divided into two basic categories: 1) locational data, indicating the size, location, and shape of a unit of data; and 2) non-locational or attribute data, giving the characteristics (soil type, land ownership, tax rates, etc.) for each unit. A third characteristic being used more and more is that of time
29


INFORMATION FLOW
Figure 1 1
30


(Dangermond, 1982). Figure 12 is a conceptual diagram of the relationship of these three elements, i.e., locational data, attribute (non-locationa1) data, time. Since locational and attribute data often change independently of one another with respect to time, management of spatial data is most effective when locational and attribute data are managed independently.
2 . Data Entry
For standard data such as maps and photos, it is necessary to convert the spatial data into machine readable form before it can be entered and placed in storage in the computer. There are a number of techniques used for this purpose, including:
a) manual digitizing,
b) automatic entry devices,
c) automatic line following techniques,
d) optical scanners,
e) manual cell encoding,
f) survey document input.
These are described below and explained diagramatically in Figure 13, page 36.
a) Manual digitizing is the most commonly used data conversion method today. The digitizer is an electromagnetic, electrostatic or electromechanical device which converts maps or other data into digital format. The digitizer consists of a flat table with an internal (x, y) matrix and a cursor. (See page 13, Figure 3.) The operator traces the cursor, or "point locator", over the map and records key points, lines, and areas which are
31


1

DATA RELATIONSHIPS
Figure 12
32


read directly into the computer. This process is quite time consuming, usually requiring a greater investment of time than all the other aspects of a GIS project combined.
b) Automatic entry devices refer to a variety of emerging technologies and instruments which directly capture and record spatial data automatically on site (referred to as "real time" data collection). These include satellites such as LANDSAT and related image processing technologies, digital photogrammetric
systems, airborn geophysical equipment and "auto surveying"
• • 1 ? devices which capture coordinates as they pass over terrain.
The main advantage of these systems is the timeliness and
accuracy of the data. However, at this-time, most of these
systems work with a relatively large scale (e.g.,1.1 acre
resolution for LANDSAT's first three satellites), making them
impractical for certain applications. This problem is being
remedied, though, as LANDSAT's most recent satellite (NASA Module
8-21980) is capable of producing spatial resolution of 30 x 30 1 3
meters.
c) Automatic line followers (ALF) are laser driven devices which are used to scan off a continuous string of coordinates associated with lines of a map. The ALF traces the lines by sensing the difference of light reflectance between the line and the background. The devices have the big advantage of time and manpower savings, but currently are uneconomical for most applications.
d) Scanning devices, similar to ALF's, automatically
33


convert printed material into a machine readable matrix form by sensing different intensities of reflected light. They also hold promise for the future, but are still in the initial stages of practical usefulness.
e) Manual cell encoding involves encoding points in a grid assigning dominant features a higher numerical value. These are written on a coding form, then keypunched on cards and fed into the computer. Used in the early stages of GIS development, this process is time-consuming and impractical today.
f) Input of survey documents is typically done through the manual key entry of x, y reference points, bearings, and distances relative to specific features desired for encoding. Features commonly encoded include ownership boundaries, roads, and buildings. These are especially useful when working with land ownership, census data, grid utility mapping, but are little used for LA analysis.
Emerging automated data entry technologies will radically and positively alter the data conversion process. Speculations are that it is only a matter of five years before these technologies are viable.^
3. Data Structure
Data structure influences the type and volume of data, the way in which they are collected, their resolution (distinction between closely related objects), validity, and ease of manipulation.
The three types of spatial or locational data used in a GIS
34


are points, lines, and areas (see Figure 14). Points are used to identify locations of features which are so small as to have no areal context (waterfalls, wells, mines, etc.). Lines are used for linear data such as streams and roads. Areas, or enclosed spaces, are the most common geographic data and include land use types, vegetation types, soils, counties, etc. The two types of areal data structures are grid and polygon.
In a grid system, (also referred to as a raster system) a uniform grid, typically square, is super-imposed over the study area, and the predominate attribute of each grid cell is entered into the system (see Figure 15). This allows certain types of manipulations which can't be done in polygon format (i.e., the matrix manipulations used with most programming languages), or which can be done more quickly in a grid format. The major disadvantage is that, because of the genera 1ization-of cells, exact locations may be lost. This kind of process is best done for very large areas where some kind of generalization is necessary or desirable, or in situations where computer memory is 1imited.
In a polygon system, the attributes are directly and fully represented*. The two types of polygon systems are polygon loops and arcs.^
Polygon loop systems represent the first generation of polygon systems. Each polygon is digitized separately and stored as an individual record. Consequently, the boundaries of adjacent polygons are digitized twice. Since it is almost impossible to exactly duplicate the line each time it is
35


TYPES OF SPATIAL DATA
Figure 14
36


digitized, an undesirable phenomenon known as slivering occurs (see Figure 16). This causes overlapping lines or unnecessary spaces between polygons.
With the arc structure, only the boundaries between polygons are digitized, and the left and right polygon labels (i.e., attribute names) are entered at the digitizing stage. Since each line is digitized only once, slivering is not a problem. For this reason, the arc structure has become the more desirable design.
Also important in the structure of a geographic information system are the location identifiers, or geo-referencing systems. Latitude and longitude and other x, y coordinate systems are the most common methods used for indicating spatial location. Geographic data can also be represented entirely without coordinates, using principles of graph theory involving topographical relationships to express relative location. This type of geo-referencing system is most often used for concepts relating to networks (streets, rivers, etc.) or when any graphing function is required. Some GISs incorporate both of these geo-
referencing systems.
>
4. Editing and Updating
Both while the data is being converted and entered in the computer, and after this digitizing process is complete, the data needs to be edited and checked. Some checks are done by the software, making this process automatic. Those systems which do offer this feature are obviously more convenient and can save a
37


’Vou^t doXa uUOwss CoprkdUtti? ddtto pdut^ ripUuL pci bX
pro dati editing CCpohdiJtior
dddc urmpkti am pr part art, add. aw aro delete node,
CfuMOrf noJLt
ToUjqcn dMd edUZnjh CltUU pOltytyn, ckautp portion, rf pohffltm, add poUjffM
fr-ti dafr, edetinr
ddtiz wuti or fauldpU colls
CAjuqr, ctUr/upiM-dics rf ieusti ^ ^odiyfo celti-
Ujldoc fydfa a*4 thuloipO. toil*
pfrlh-jJ dati idotiHsj
add, UuoKff, dehlo, reptau- aAxrorti,u
fli/vO 5ctj of Ckxsrudo*
dddc, add, rudact, eA*r*dC~ jturyp tiJCnd far %trt«Q
INTERACTIVE EDITING FEATURES
Figure 17
38


tremendous amount of time. Some common elements of initial checking process are:
a) plotting or printing out digitized data for visual editing; including checking it with the original for discrepancies,
b) checking for polygon closure,
c) checking with surrounding data to make sure they match (referred to as edge match),
d) deleting duplicate or unnecessary lines (slivers) or points,
e) checking polygon labels.
The updating frequency depends somewhat on the data type. Certain types of data, such as geologic features, are fairly stable, and if entered accurately needn't be changed for years. Other data, such as vegetation or demographic dat£, .change frequently. Editing features vary from system to system. Some basic interactive editing features to look for are listed in Figure 17.
5. Storage of Data on Tape or Disk
As data are entered into the computer through means of manual or automatic digitizing techniques, the Data Base Management System (DBMS) of the GIS software takes care of storing the data on magnetic tape or disk. In the process, it sorts the data and files it in such a way as to allow for the greatest efficiency when later searching for specific data types.
If the user is fortunate enough to have located a source for
39


pertinent data which is already in digital form, this is the stage at which the user can transfer the data to his/her system, re-formatting the data as necessary.
6. Data Retrieval, Analysis, and Manipulation
Computer analysis is, of course, the primary function of a GIS. This section describes the various analytic and data processing functions that can be performed by a GIS. In terms of minimum data function, there is no standard definition of a GIS. Tomlinson suggests that if the manipulations of search, measurement, and comparison are not needed, then a GIS is not needed.^ Most GIS functions include combinations of these three components.
Searching is the capability to read an entire file and retrieve any specified set of data by location, attribute, or specified attribute value.
Measurement functions are shown in Figure 18. These include calculations of area, perimeter, distance, volume, and direction.
Comparison is probably the most powerful capability of a GIS for LAs. It uses descriptive or location data to determine relationships based on criteria from one.or both types of data. The most common comparative techniques are overlaying and compositing. (The technique of overlaying, incidentally, was pioneered by well-known landscape architect Ian McHarg.) Overlays are very useful in building suitability maps. Relationships of maps or attributes can be described quantitatively and comprehensibly in a matter of minutes.
40


MEASUREMENT FUNCTIONS
Figure 18
DIGITAL TERRAIN ANALYSIS FUNCTIONS
Figure 19
41


(Section B in this chapter outlines some specific overlay examples ) .
The various digital terrain analysis functions outlined in Figure 19 and described below should be of special interest to LAs. These functions are computed from a digital elevation model.
Visibility -- primarily used in conjunction with determining what can be seen and not seen from a given x, y and z coordinate. Output is in the form of a grid of values indicating seen and non-seen values.
Sun Intensity/S1ope/Aspect — primarily involves the calculation of slope aspect, and in some cases, the relationship of this slope aspect to solar radiation or sun intensity.
Interpolation/Contouring -- typically involves the ability to take random or regularly spaced data and generate a grid or other structured model framework on which automated contouring can occur. Based on this interpolated model, a contouring program actually computes and subsequently draws out the contour lines.
Watershed Computation -- involves the computation of the watershed boundaries, drainages, topographic valleys and peaks. In some cases, these data are used to compute surface flow.
Visual Display -- consists primarily of cross sections and three-dimensional views.
Other GIS manipulations of interest to LAs are listed briefly below. Most of these represent combinations or
42


variations of the basic functions already described. It should be noted that there are many more GIS functions than those listed here; these are meant to be representative overall, with an emphasis on those which could be most useful to landscape architects.
Graphic query
Conversion -
Sealing/Pro-portional
Buffering -
Windowing -
Corridor Delineation -
Boolean Attribute Retrieval -
define an area of a map on the terminal and ask to see everything about it.
convert data from small units such as Douglas Fir to larger units such as evergreen vegetation.
size of output from 3' by 3' to 8-1/2" by 11" or scale of data from 1" - 20' to metric
Delineate a 20' zone around a facility or road and analyze and display all the cost alternatives in the zone.
Zoom in to a small area or step back for the big picture.
Based on a number of criteria, find a corridor for a road, powerline, or oth.er linear feature.
Determine all of the polygons of a certain type of soil which have an area of greater than 50 acres.
Statistical Summary -
For the above (boolean) example, generate a statistical summary of the polygon acreages, perimeters and their totals.
7. Output Vormats and Output Devices
Output products vary in format and are dependent upon software, hardware, and user preference. They include 2-D, 3-D,
and 4-D maps, graphs, charts, and reports. Various GIS output are reproduced in Figure 20. Options such as color and shading also vary from system to system and may be dependent upon accompanying hardware.
43


GIS OUTPUT
Figure 20


The attention given to the map in a GIS should not overshadow the importance of report generation. Beyond simple tabular listings, a report generating program can provide:
a) summaries of system status (tables of attribute and locational data, size of data base, use statistics, update information, etc.),
b) statistical tables,
c ) user guides,
d) reports for presentation.
Output devices are many: line printers, fixed carriage typewriters, drum and flatbed plotters, electrostatic printers, Cathode Ray Tube (CRT) terminals, computer output on microfilm (COM), color xerography, and color slides. Of these, the most commonly used are the CRT and the plotter.
The CRT (Figure 3) is the TV-like screen which displays data as it is being manipulated. "Interactive" or graphics software allows the user to specify changes -to a screen image and will then make the changes instantaneously; this feature is considerably more convenient than "passive" graphics software which would have to take the time and effort to redraw the entire image.A CRT can be interfaced with printers or plotters for a permanent image.
The plotter (see Figure 3) produces a permanent image on paper, acetate or film. A number of drafting/drawing pens can be used to produce the image: felt tip pens, technical ink pens, ball point pens, scribing tools for directly producing photographic negatives, or a point light source for exposing
45


photographic negatives.
In general, line printers and fixed carriage typewriters are less expensive than plotters, but produce an inferior output, generally unsuitable for presentation maps.*®
Computer output on microfilm (COM) produces a fine resolution output which is convenient for viewing and storing maps which can be enlarged. Color slides are relatively new options for output which also provide fine resolution output.
The two main issues in considering map output format are: 1) for what will the output be used - for working level or for presentation, and 2) what degree of resolution or level of accuracy is needed. If presentation quality maps and a high degree of accuracy is required, the plotter is the obvious choice. If these are not needed, the line printer or fixed carriage typewriter can be relatively inexpensive alternatives.
D. Classification of Systems
As mentioned earlier, GIS uses cross over many professions and applications (see Figure 1). Though many of the functions overlap, GISs in general cover a broad spectrum of focuses. GISs vary in design, orientation or focus, and thus in their overall capabilities. Consequently, the degree to which individual systems are appropriate for the LA profession or for specific applications varies. This section classifies GISs according to 1) orientation or focus, and 2) size of the accompanying hardware.
46


1• Orientation of the System
Dangermond classifies GISs according to the following categories based on the purpose of their initial design.The first three are of most interest to LAs.
Generalized Thematic and Statistical Mapping Systems -Typically on mini or mainframe computers, these are the systems most commonly utilized by LAs. Used for natural resource management, forest inventories, vegetation, geology, soils, census mapping, environmental planning and assessment, etc.
CAD/CAM Adaptations - Typically a mini-computer CAD/CAM system for engineering applications such as photogrammetry, topographic basemaps, road engineering, and utilities. These systems vary in the amount and complexity of GIS analysis which is possible. Many of them are primarily graphic systems, and consequently lack some of the analytic and location functions expected of a full-fledged geographic information system.^ Their main limitation for landscape architects involves overlay and composite work. Graphic overlay and composite information can be generated, but analytic comparisons and relationships in terms of specific coordinates are not usually possible. Also, if
N
location identifiers are not used, important functions such as "visibility" cannot be performed. Their main advantage for LAs is their generally superior graphic output. Because of their focus on the design and visual elements, there is generally more focus placed on the quality and quantity of output options. It has been suggested that the functions and focuses of GISs and CAD


systems are merging and that these once separate and distinct technologies are now taking advantage of research and development in each other's field.22 it is important to note the distinction between the two at this point, however, to assure that when the full GIS capabilities are needed, they are guaranteed.
Image Processing Systems - Usually associated with processing LANDSAT and related satellite image data, these are also used by LAs in large scale regional projects.
Property or Parcel Information Systems - Typically a mainframe based data base management system to handle attributes associated with the land parcel.
Geographic Base File Systems - Most often associated with street networks and the aerial units which they define (e.g., the DIME system used by the U.S. Census Bureau).
Bibliographic Systems - Catalogues a variety of bibliographic data sets about geographic documents.
The important thing to note here is that different functions support different levels of accuracy, different types of data management, different spatial identification techniques, and different approaches to information display. There is some overlap in system types, but knowledge of the basic uses for which a GIS was initially designed can provide a good clue to its emphasis and can offer help in judging its appropriateness for a particular project. For example, if the main purpose for
48


employing a GIS is to do a statistical analysis of areas and relationships (e.g., for preparing environmental impact statements) a GIS which has been adapted from a CAD/CAM system may not have all the capabilities needed. If, however, graphical overlays will be a major emphasis with only minimal specific analysis needed, a CAD/CAM adaptation may be suitable.
2. Size of the System
GISs may also be classified according to the size of the computer used with the GIS. This determines speed, the size of the data base which can be effectively handled, and resolution. Early GISs could only be run on the largest class of computers, the mainframes. As the mid-sized computers, the minis, grew in power, and as the GIS software became more efficient, GISs were developed which could handle large data bases on the minicomputer. Today, GISs are used even on micro-computers, although with some degree of trade-off in accuracy and resolution.
IMGRID, a GIS developed at Harvard University in the mid 1970s is one such system.^ IMGRID has strong analytic and modeling capabilities and has been used extensively for site selection. Researchers found that the mainframe system could be adapted for use with the mini and ultimately with the microcomputer by reducing the data from polygon (curvi1enear) to raster (grid) form. There was a significant but acceptable degree of loss in processing speed and resolution, yet there was also a radical reduction in system cost.
49


The significant reduction in the cost of microcomputer geographic information system capabilities as compared with large computer systems provides an opportunity for an expanded and diverse user community including such small businesses as landscape architecture, planning, engineering firms, and universities. 1
For LAs, the main considerations in this selection process are site size, scale of analysis, and degree of resolution necessary on any given project. MacDougall suggests that "most regional landscape studies will probably continue to be done on mainframes and mini-computers until there are very significant increases in the speed of micro-computers." Nevertheless, it is clear that certain kinds of analyses can be handled quite effectively by the smaller computers accessible to small LA firms.
E. Considerations for Landscape Architects
Because of the unique orientation of LA work, this section summarizes the basic considerations about GISs for the landscape architecture profession by examining issues of particular concern to the profession. These issues include: advantages and disadvantages of GISs for landscape architects; data scale and accuracy; GIS costs, and finally, our current methods of work in relation to^emerging computer technologies.
1. Advantages
The main advantages of geographic information systems for the profession include:
1 . The designer or planner can consider many more factors because the GIS can help put the information (or combinations of information) in a more comprehensible
50


form. Easily accessible information includes not only environmental data, but also socio-economic and administrative information.
2. Rapid and repeated conceptual models or development scenarios can be investigated and tested against criteria established by the LA. This increases the likelihood that the optimal alternative will be identified and also allows for a more clearly defined process with quantitative results.
a. It is also possible to test models with regard to
changes over time.
b. Client participation can be encouraged in the decision making process through the testing of various models.
3. It is possible to do analyses that would be impractical or inefficient to do manually, (e.g., in-depth runoff calculations for very large areas). Use of a GIS can thus increase the accuracy and credibility of overall results.
4. Analysis data, design and management plans can be maintained in a physically compact format (i.e., the magnetic file). This reduces "paper shuffling" of maps at different scales, reports, tables, and other text. Essentially, an efficient work space/filing system is built into the system.
5. It is possible to quickly and efficiently change scales and to work on portions of a project in more or in less
51


detail.
6. Editing and updating of the data base is fast and efficient. The system has built in checks for common errors.
7. Among the most useful analytic manipulations which can be done quickly and efficiently are:
terrain analysis (slope, aspect, cut-fill calculations), sun/shade analysis, overlays,
3-D modeling,
view shed delineations,
area and frequency analysis,
drainage maps indicating direction of flow, sink holes and ridge tops.
8. More effective presentation can be made through the use of sophisticated graphics output, e.g., three dimensional perspectives showing what the site will look like after development.
9. The productivity of individuals or groups using GIS maps is increased because of the greater consistency, versatility, and accuracy of a GIS.
2. Disadvantages
There are also a few disadvantages to using a GIS which need to be considered by landscape architects. Some of these are:
1. High initial hardware and software cost. The prices
52


vary depending on the size of the system, but even a small, basic system will cost around $15,000. (See "cost" later in this section for further discussion.)
2. Time and money to convert data not already in digital form. For this reason, a GIS is most advantageous if the data are to be used more than once. Most landscape architectural offices, however, work on sites on a one time only basis. Until scanning and automatic entry technology advance to the point of substantially reducing costs, then, smaller one-time projects may be cost effective only if the data are available in digital form, or if limited data are used. Certain kinds of data require much more time to input and may not be needed for every project. For example, soil data are some of the most time consuming to.digitize. Fortunately, in many areas, Soil Conservation Survey (SCS) information is already available in digital form. For those areas where it is not available, however, all the soil data may not be required for every project. The analysis may only require generalized categories, which would allow ditigizing of only larger categories of data; for example, clay soils, sandy soils, etc., instead of the more specific categories mapped by SCS. This could save a great deal of time and expense, and still be of sufficient accuracy and detail for many types of analyses.
3. Landscape architectural orientation is * not built into
53


many GISs. Many GIS companies are just beginning to realize the market for the design professions. In the past GISs have tended to be built and designed for whomever could afford to fund the necessary research and development. Consequently, most initial systems were developed for utilities, engineering and geophysical exploration applications. Although these can, and have been, adapted for LA applications, it is important to understand the possibile uses of GISs for LAs so that we are able to communicate our needs to GIS companies as they develop and modify their systems to better suit our purposes.
3. Scale
Most landscape architects who are familiar ..with GIS technology consider them for use only at the regional scale. In the statewide survey of the profession which was conducted for this project, for example, 81% of those who had some understanding of GISs considered them valuable only for regional studies. While this has been true in the past because of the cost of mainframe and minicomputer based systems, today LAs are using the technology on smaller systems for projects of varying scales. S.J. Camarata, a landscape architect and president of IRIS International, a GIS company specializing in micro-computer software for LAs, says that he now employs a GIS for all of his projects, regardless of scale. He notes two projects in particular for which he used ERDAS 400, a micro-computer GIS.
54


One project was a nationwide study for the Nile River Valley. The scale and resolution of the project were on a gross scale typical of projects this size. All analyses were done on the microcomputer GIS, though most manipulations and mapping were performed overnight. In contrast to this project, Camarata discussed smaller site specific applications using the same microcomputer GIS. For these projects, data bases were limited to information available digitally or which could be digitized fairly quickly.^4 Primary manipulations included topographic and elevation data for modelling runoff and slope analysis.
4. Accuracy
One of the advantages discussed earlier in this section was that GISs can provide quantitative results, ensuring greater accuracy in the overall analysis. While this is certainly a valuable feature, it is important to understand (and to be constantly reminded) that quantitative results can only be as accurate as the accuracy of the original data entered.
For this reason, determining the appropriate scale and detail at the outset can ensure 1) that the data are appropriate for the levM of analysis needed, and 2) that time is not wasted digitizing unnecessary data.
This problem was exemplified in the NWCCOG project which was discussed in Section B of this chapter. On the site-specific analysis for studying development suitability, the set of soil data was one of the main factors in the site selection equation. Unfortunately, instead of collecting all the pertinent data, a
5 5


decision was made to use readily available data from the Soil Conservation Service (SCS). Not only were the SCS maps insufficient for this analysis because of their 1:20,000 scale, they also failed to provide the necessary soil depth profile. Because of this lack of accuracy, the results were worse than useless - they were misleading. The consequences for this section of the NWCCOG project were disastrous.
5. Cost
GIS software is available from private companies, some public entities, and some academic institutions. Most software from public entities (e.g., BLM, U.S. Forest, U.S. Fish and Wildlife Service) is available either free of charge or at minimal cost, while academic institutions provide software on a proprietary arrangement basis. Although the initial .software cost is lower, usually support services are not provided, and the software is generally less versatile, having been developed for specific "in-house" type projects.^5
Frequently commercial GIS software is provided in one or more modules, so that a client need not acquire all the software that an organization offers. Such modules typically start at around $10,000. An average GIS costs approximately $100,000 for all needed software modules. The hardware will typically cost the same or more.^ GIS micro-computer turnkey systems (i.e., systems with all the necessary software and hardware included) range from approximately $15,000 to $75,000 in price.
At present, there are GIS systems available at the "low end"
56


($15,000 to $75,000) and at the "high end" ($150,000 to $1,000,000) but not many in between.^
6. Rethinking Our Old Methods
It has been suggested in this paper that computers in general and geographic information systems in particular, have the potential to restructure the form and mode of LA practice. Discussions so far have emphasized improvements in terms of accuracy and efficiency. While this is an important change, it is only a part of the potential for change and growth.
Now, more than ever, it is essential to step back and look at our old ways of doing things. Before we automatically transfer those old methods onto computer technology, we must ask if those methods are still valid. The technology offers us tremendous opportunities to view and to question the.possible consequences of our decisions before taking decisive action. This has enormous implications for the LA profession, and the potential to radically alter the way we go about our business.
In examining trends in the evolution of GIS technology, a recent article in the Harvard Library of Computer Graphics asserts that? "the rapid evolution of GIS technology has not been matched by a corresponding increase in our understanding of how natural systems work. Thus, as in other areas, technology is probably outstripping our ability to apply it intelligently; our reach exceeds our grasp considerably."^ It is up to us, then, to apply the technology intelligently, to remain cognizant of the fact that the computer does not perform feats of magic. The
57


quality of data entered and the methods of evaluation and analysis used will always be reflected in the quality of results and output.
N
58


CHAPTER III
GIS EVALUATION CRITERIA FOR LAs
This section outlines a basic set of criteria for evaluating GISs in relation to the unique set of requirements of LAs. Since software is the primary consideration, in selecting a GIS, the general thrust of this section is on the selection of GIS software, but a few comments are offered about basic hardware considerations as well.
There are a host of GISs and related mapping software from which to choose. In order to select the system most appropriate for your purposes as a landscape architect, a four step process is suggested here to help narrow the range of GISs to .consider. The process as outlined in Figure 21 is to 1) assess your specific needs and define the tasks for which the GIS will be used, 2) determine the appropriate size and orientation of the system, 3) evaluate the reliability of the GIS vendor, and 4) appraise the GIS for ease of use and flexibility. Some considerations regarding system costs are also discussed briefly in the final portion of this chapter.
A. Establish Your Needs and Determine Your Requirements
Evaluating any computer system for a specific situation requires a thorough understanding of how things are currently done. The organizational structure, the physical work settings,
59


â–  DEFINE YOUR NEEDS
|f Silt /TYPES OP rrVJECTS USED
•iucHotujJ fX4JVi ■ rtqlofvU icwiotiw pWc. • site' epaajAc. 4«V«jn
sire op ocrrpur
•»>t >cu •x • Ho K «>
- -' *
OUAUTY opavm>r
* prc^c ntafatl

DETERMINE
SIZE AND ORIENTATION OF THE SYSTEM
(MtimuM of rue sre^tA
‘ +hfciW4i£, • CM>{ CM\ oAa^jMok.
5111 OF TH£-
WOKYK •Micro latsei • MiVu l&$ed
â–  CHECK THE RELIABILITY OF THE VENDOR
• Okwwtcc -me
• &prvj^
• T^€o • ofcr*-' o^«>- lufvr
• -ptueMctH, -e • 4 PEvetofM^ur
EVALUATE THE GIS FOR EASE OF USE AND FLEXIBILITY
< Tire e^‘5Hfw i^uc* ydva owu pp^eer
♦ MgwU ttust&i <
» Kotep-Aorive ?
• dpcvm&wicu njRrrrBW r
• K«ft PHfeWPlY <"
♦ OoOew ewe mmu?
- wfiw-eie -w&Zr K>-Vlrr*c/& W lu^euACtence ce flie TW«Mi»JAt,r
CRITERIA FOR EVALUATING A GIS
Figure 2 1
60


and the type and amount of work done all need to be examined in detail. How will the computer change this structure? Will it expand or eliminate the work that is handled in the setting? A close examination of these issues and an establishment of goals and designs for the future will make the transition to computer use more effective and more friendly. This should be the first step in the GIS selection process. Examine how a GIS will fit into your particular situation. How will the system emulate or change the way things are currently done? How many people will use the system at one time?
Having gained a general understanding of geographic information systems, define the purposes for which the GIS will be employed. List the basic manipulative functions which will be used and the order of their relative importance. Determine the size and types of projects which will be studied, the- range of scales which will be used, the type and amount of data to be analyzed, and the degree of resolution and accuracy required. How quickly do you need results? What is the quality and range of graphic output that you require? Answering these questions and establishing priorities will serve to focus your search.
B. Determine Appropriate Size and Orientation of the System
The next step in narrowing the search for an appropriate GIS is to determine the appropriate size and orientation of the system. This is based on the discussion of classifications of systems in Chapter II.
61


1. Size of the System
In general, the more demanding the requirements, of the GIS the larger and more powerful it will need to be. This is particularly true if speed of data access and manipulation is important, if large-scale projects will be undertaken, or if a large number of people will use the system at one time.
As was discussed in the introduction to Chapter II, many of the functions of a CAD system and those of a GIS overlap. For this reason the considerations for purchasing CAD systems and GISs are in many respects quite similar. In his article "Wary Buyer's Guide to CAD System Demonstrations," Tim Varner discusses the issue of selecting the appropriate sized system for your needs. In deciding between a mainframe based system and a less expensive mini based system, he suggests that "there;, are gray areas where the capabilities of the differently priced systems will overlap. Therefore, what you may be paying for in a more expensive system is the speed of operation. Keep this in mind when defining your needs and priorities, to make sure the extra speed justifies the extra expense." Varner also maintains that along with increased speed of the mainframe-based systems come increased price and often increased problems. A system crash on a mainframe computer, he notes, could result in several days of
O Q
down time for all of its users.
A mainframe-based system may be justifiable only if you require high speed transactions, have a large number of users (e.g., over 25 at once)^, or if you have a mainframe computer
62


already. In most cases, a mini-computer GIS will have comparable capabilities to those of a similar mainframe based GIS.
The main advantage of the micro-computer GIS is price. If your budget is tight and your needs limited and specific, a micro-computer system might be the answer. There are currently a number of micro-computer GISs designed for LA work. IRIS International, Inc., for example, has geared their software to the needs of the design and planning profession.^ It has been suggested that at least 2-4 megabytes of memory are generally needed for micro-computer GIS analysis.22 For more discussion on micro-computer GISs, see Chapter II, Section D.
2. Orientation of the System
Having selected the size of the system, it is helpful at this point to narrow the selection further by determining the orientation of the system which is appropriate for your needs. Using the classification system presented, it was suggested that the three orientations most relevant to LA work are 1) generalized thematic and statistical mapping systems, 2) CAD/CAM adaptations, and 3) image processing systems. To determine which of these systems would be best suited to your purposes, review your needs as established in the first step of this process. If, for example, it was determined that satellite data will be the primary data source, reviewing primarily image processing systems will serve to focus your search.
63


c.
Evaluate Reliability of the Vendor
A criterion of major importance for evaluating GISs is reliability of the vendor. Some basic questions to answer in this regard are:
1) Is the software guaranteed?
2) Are software updates provided? Free or at a price?
3) Is the training of personnel included in the purchase price? Varner asserts that "an intensive training class can reduce - by as much as three weeks - the time required for . . . operators to be productive on the system." He advises wariness of vendors who claim such user-friendly systems that the company provides only two or three days of classroom instruction and a toll free number to call if problems arise.^
4) How accessible are the vendors to give advice and instructions and to answer questions?
5) Who are the current users and what has been their experience with the system and with the vendors?
6) Does the vendor write the software or is it purchased ffom a second party?^
7) Can you receive hardware and software service locally and at what cost? Robert Puterski, senior cartographer for Colorado Department of Local Affairs, claims that local servicing can mean the difference between 1-3 days for local servicing versus weeks or even months if equipment must be sent back to the manufacturer. The
64


industry standard for such services ranges from .75 to 1.5 percent of the purchase price of the equipment per month.3 5
8) How financially stable is the vendor? Does the company have the personnel and finances necessary to continue the development and enhancement of its product? How large is its research and development staff? How large is the service and support organization?^
D. Appraise the GIS for Ease of Use and Flexibility
This section provides a list of GIS features which are desirable because they 1) make the system easier to use and/or 2) allow for flexibility so that the system can be updated to suit your needs and adapted for new developments in GIS technology. These factors probably most easily discerned at a GIS demonstration.
When arranging a demonstration, it is important to communicate your needs to the vendor so that s/he can tailor the demonstration to meet your specific objectives. If possible, arrange to bring in some of your own work with which to try the system. This is a good way to observe the system being used "fresh." Canned presentations can be misleading.37 Some features to evaluate during the demonstration are:
1) How user-friendly is the system? Look for techniques or operations which seem unusually complicated or cumbersome. How often does the operator use the
65


vendors?
How flexible or expandable is the system?
39
E. Cost Considerations
One way to ease the high up-front cost is to purchase the software in modules, purchasing only a core system at first and adding on when additional funds become available.
For LAs who will use the GIS for projects where the data will be used only once, data base updating and editing features may not be important.40 It may be desirable in this case to purchase a system without this expensive feature of the Data Base Management System (DBMS); this can sufficiently reduce required memory, cost, and overhead.
The four step process outlined in this chapter is meant to clarify the most important issues concerning the evaluation and selection of a geographic information system for a given purpose or setting. Following each step in the process and answering the questions listed, it is hoped, will narrow this selection to a few of the most appropriate GISs in each case. These systems can then be evaluated against one another based on the specific needs and priorities established by the potential users. The final evaluation should examine each system based on how well it can be tailored to the needs of the users, how close it comes to the designated budget, and to what extent it will grow and adapt to the changing needs of the users.
67


CHAPTER IV
RECOMMENDATIONS TO UCD LA PROGRAM
The initial recommendations made here are modeled after the previous Chapter, "Criteria for Evaluating GISs". First, the needs and requirements of the LA program at UCD are discussed. Appropriate GIS classifications for UCD are then determined, and And, suggestions are offered on how to incorporate GIS technology into the LA program curriculum. Finally, some logistics concerning staff and training are discussed and options for cooperating with public and private agencies are explored.
A. Needs/Requirements of the UCD LA Program-.-
While the primary purpose of using GIS technology is to aid in the design and analysis process, a second purpose exists in the unique environment of an educational institution. This second purpose is to expose students and professionals to the capabilities of GIS technology and to provide training so that it becomes a useful and comfortable tool.
Since the LA curriculum at UCD encompasses the wide range of aspects of the LA profession in general, the needs of the program in the area of GIS technology are quite diverse. All the GIS functions suggested in the previous chapter as most advantageous
68


to LAs in general, then, would be useful throughout the curriculum. Basic modeling functions, for example, would be useful early in the design sequence, while certain functions such as overlay would be most useful later in LA 601 (Design IV) during regional design. (More suggestions regarding incorporation into the curriculum are discussed later in this section.)
The size of projects and scales vary from smaller site specific to large regional projects. For smaller projects and site specific analysis, the higher resolution polygon data structure would be required, while for regional projects the more efficient grid structure would often be necessary to reduce memory requirements and to increase speed.
Output sizes are a function of the use to which it will be put. Since most projects are used for presentations and some are used for publication or photographing, the larger input/ output devices are most desirable (e.g., a large digitizing table and a drum plotter).
The number of users who would be on the system at one time is the most difficult factor to determine. The system could ultimately be used by LA students and professionals, as well as
N
by other students in other divisions of the college. As suggested in the "Classifications of Systems" chapter, this is a crucial factor in determining the size of the system. This decision should be based on 1) amount of space available, 2) the commitment (money and staff) on the part of the university, and 3) interest in GIS technology from other divisions within the


College of Design and Planning.
As shown in Figure 21, the last consideration for assessing the needs of the LA Program is the speed required of the system. This is another area which should be defined by the program. The advantages of a higher speed system would be a faster learning curve, a lower frustration level, and increased likelihood that students would consider GIS technology as a viable tool for use on projects in a professional setting. But if the university's commitment (i.e., funding and staff) are minimal, and a micro based GIS is thus acquired, this handicap could be minimized through the careful structuring of projects and assignments.
B. Appropriate System Size and Orientation for UCD
The next step suggested in evaluating a GIS for UCD's purposes is to classify the system in terms of it-s overall orientation, and in terms of its size.
1. Orientation of the System
The systems which would be best suited to the LA Program
would probably be those most commonly used by LAs in general, \
i.e., generalized thematic and statistical mapping systems. These are the systems most often used for regional design and analysis, but they can be used for a number of other purposes as well.
While not as versatile overall for the needs of the program, GIS CAD/CAM adaptations could nonetheless be integrated into
70


parts of the curriculum if it is determined that purchase of a GIS per se is not feasible. This would be desirable if the College of Design and Planning purchases only a CAD system with mapping software as an adjunct component. This type of system would be most beneficial for use in first year design classes for modeling and site analysis and for some kinds of engineering work, yet would be of limited use for regional analysis and design.
2. Size of the System
The process of establishing needs and determining GIS requirements will generally serve to define the size of the GIS which will be most appropriate. The particular requirements which most affect what size GIS is needed are 1) number of users on the system at one time, 2) speed of transactions, and 3) budget. Since these three factors have yet to be defined by UCD, a single appropriate GIS size category cannot be established here. This section will therefore discuss advantages and disadvantages of the three GIS size options for UCD.
Chapter III suggests that mainframe GISs may provide some increase in speed of transactions, but that their additional expense and increased problems often do not warrant their selection over minicomputer GISs with the same fundamental capabilities. A mainframe GIS would be impractical for use at UCD for these reasons also because its larger size could not likely be integrated into the limited computer space currently available in the College of Design and Planning. Several space
71


alternatives are currently under review, but it is probable that space will be at a premium until the new UCD facility is built on the Auraria campus.
Minicomputer GISs with the same basic capabilities as mainframe GISs, but with higher resolution, greater speed, and more available memory than micro-based systems would likely be the most practical and versatile for our purposes at UCD. They could be incorporated into many aspects of the LA curriculum and could be used by other divisions in the College as well. Because of this diversity, they would be better able to grow with the changing needs of the College of Design and Planning, would allow for more experimentation, and would be more feasible as a resource for use by outside professionals.
The biggest disadvantage is their high cost in relation to
microcomputer GISs. This would require a larger university
commitment in terms of both money and staff. While various
funding options were previously suggested, these would be worth
pursuing only if the system is used at something approaching its
full potential. This would require staff to teach introductory
computer courses, to train students on GIS use, and to organize \
and supervise a variety of GIS projects.
Though not as versatile as a minicomputer GIS overall, a micro-computer GIS might be a good introductory system for UCD. While a microcomputer system would be considerably less versatile and unable to handle as many users at one time, it has the advantage of being less expensive and more compact. These
72


factors increase the likelihood that a GIS could be implemented at UCD more quickly and used sooner. A micro-based system cound be used for smaller projects on a limited basis as a way for students and professors to become conversant with the basics of GIS capabilities and the structure of GIS projects. A phasing plan might be adopted whereby a smaller, less expensive micro GIS would be purchased initially to familiarize students with some basic modeling and analysis manipulations. Then, as staff, available space, funds, and needs grow, a larger minicomputer GIS could be purchased. The micro GIS would be maintained for introductory courses and used for first year projects (see "Incorporation Into the Curriculum"), the minicomputer system would then be used only by those students and professionals with GIS experience and for larger projects requiring more users at once, or for thesis projects where higher resolution would be necessary for in-depth analysis. UCD students would then be trained to use both systems and would understand their potential for the variety of LA positions available to graduates of UCD (i.e., small LA offices, planning offices, public agencies, etc. ) .
This f^inal option of purchasing a microcomputer GIS now and expanding to a larger minicomputer GIS later appears to be the best option for UCD. First, it ensures earlier implementation and GIS use and second, it offers the broadest and most well rounded program overall, including as it does both microcomputer and minicomputer GIS applications.
73


3. Cooperation with Public and Private Agencies
To increase the visibility of students' GIS skills and to enhance UCD's image, it would be advantageous for the LA program to encourage cooperation with area professionals. Two ways of structuring this cooperation could also help raise the necessary funds to purchase a GIS. The first strategy has been suggested by UCD's architecture and interior design divisions. This was a plan devised for soliciting donations for the purchase of a CAD system for the College. In this strategy LA firms could be solicited to donate money to the program's GIS fund. Since the university has tax free status, the donations would be a tax advantage for these firms. In return for their donation, the firms would be offered time on the computer, perhaps pro rated for the amoung of the donation. This time might be limited to hours or times of year when UCD student demand for GIS use is low.
The second approach to encourage professional support is by way of offering short seminars geared to the needs of these public and private firms. These could be held during summer months or between semesters when GIS use by UCD students would be low. So m^e seminars which might be considered are 1) an
introduction to GIS technology; what it is, how it is used, some special features of importance to landscape architects, 2) selecting a GIS for a specific application or office; what to consider, what to look for in purchasing a GIS, and 3) advanced GIS training.
For agencies in the area that already have a GIS (i.e.,
74


several county planning and federal governmental agencies), cooperation could be encouraged in another way -- sharing data bases and providing projects for students to work on. The agencies would benefit from having input on their projects while the university would benefit from the agency support and student exposure.
If extensive outside exchange is desired or anticipated, it would be beneficial for UCD to have a GIS which is being used by a number of agencies. UCD students would be trained on this system, have an understanding of its specific features. Interchange would be smoother and hiring UCD graduates would become particularly attractive to these agencies.
4. Incorporation Into the Curriculum
Many of the GIS applications listed in Chapter I, it has been noted, are related to the work of landscape architects. These aspects of the profession are explored in the LA program at UCD and could be integrated into the existing curriculum as GIS projects. A GIS could be used in the first year design classes for modeling and digital terrain analyses, in the second year design classes and the ecological analysis class for in-depth analysis and suitability studies, and in the engineering courses for predicting storm run off, calculating cut and fill, etc. Various statistical and predictive analyses could be explored in Research Methods, in theses, or in students' electives.
In order to be integrated into the curriculum efficiently, some basic GIS training would first be necessary. It is
75


suggested that the college of Design and Planning offer an introductory computer class which all students would be required to take in their first year. Such a course could 1) introduce students without computer experience to some basics of operation, and encourage good user habits, 2) introduce students to the computer facilities at UCD, including all the software packages which could be used for other classes, and 3) improve the general computer literacy of the Design and Planning students.
CAD and GIS training could be included as a portion of the course so that students would be ready to work on the system in the second semester of their first year. In that semester, one site could be digitized either individually or as a class, and could be used for several classes: graphics (3-D modeling and rotation), design (view shed, sun-shade, slope, and other digital terrain analyses, "what if" queries, modeling), and engineering (cut-fill, storm run off calculations, contour interpolation, etc.).
Having students share the digitizing efforts, and using the same site and same data base for several classes would eliminate much of the digitizing drudgery and thus would increase the
N
learning curve. More time would be spent exploring GIS capabilities.
In the second year, other data input options could be explored (i.e., photogrammetry, LANDSAT data, digital data bases from area agencies, etc.), and more sophisticated analyses could be investigated by the students. GIS projects would also be
76


conducive to interdisciplinary work if all Design and Planning students were given the GIS training suggested for the introductory computer course.
77


CHAPTER V
SUMMARY AND CONCLUSIONS
Early GIS technology was primarily designed for and utilized by engineering, utilities, and geophysical companies. The applications today have expanded so that users include nearly all professions which deal with geographically based data. The technology has advanced from primarily mapping functions with crude output to include sophisticated search and retrieval, composite and comparison analysis with a variety of readable color output.
Clearly, the range of opportunities for the landscape architecture profession to utilize this technology is .vast. Yet a statewide survey of the profession (see Appendix A) shows that only 8% of the respondents fully understand GIS uses and
capabilities for landscape architects. A full 85% of all
/ /
respondents though, show interest in learning more about them. It appears that the technology will continue to be under-utilized until LAs have at least a basic understanding of geographic information systems.
This paper has attempted to define GIS technology and its importance to the landscape architecture profession. GIS projects, and the classifications and workings of GISs were discussed, providing an overview of the basics of GIS technology.
78


From this understanding, then, considerations specific to LAs were developed, and criteria for evaluating GISs for specific applications and needs were established.
While a basic understanding of the technology relative to the profession is an important beginning, real strides will be made only when the technology it utilized by LAs in their everyday design and analysis process. This change will happen only when someone within the profession shows the usefulness of GISs for a variety of LA applications. It has been suggested in this paper that UCD can help affect this change and lead the way in promoting GIS use within the profession in this state.
For this reason, this project, also provides some general recommendations to UCD for implementing and utilizing the technology in the LA curriculum. Much needs to be done to fully implement a geographic information system into UCD's College of Design and Planning, yet it is hoped that this project will provide an impetus to begin this process. As a major landscape architecture university in Colorado, UCD must assume this leadership role both for the sake of UCD students who will be competing ^or jobs nationwide as well as for the sake of the LA community in Colorado who need this valuable tool in facing major decisions concerning growth and change along the Front Range.
The rapid changes in the computer industry are occurring in the GIS industry equally rapidly. Many of these changes will help make GISs even more viable for landscape architects. Some of the more notable changes are:
79


1) The falling costs of hardware and the explosion of affordable microcomputers. More GIS software is being written for smaller, more powerful computers making the technology more affordable to more LAs.
2) Increases in interactive graphics capabilities. These will serve to make GISs more accessible and easier to use.
3) Increase in available digital data bases and new options for converting data into digital form. These will decrease the heavy up-front time and money spent digitizing.
4) Real time data collection. With improvements in sophisticated satellite data gathering techniques, we can work with data which is being gathered while we work.
5) Worldwide networking of information. As data bases become more accessible and widespread, networking of information among areas and professions will occur, increasing interaction among professions and furthering understanding of our world.
These changes, coupled with burgeoning interest within the profession have the potential for profound changes in landscape architecture in the next few years. Clearly, the computer
N
industry is affecting the way we work, the way we communicate, even the way we look at the world. It is imperative that we as landscape architects become aware of these changes and become involved in how the technology affects our work, as it inevitably will. GISs are one computer tool which can help "increase the productivity, credibility and effectiveness of landscape
80


architects"41 it is time for us to become informed and actively involved in affecting the direction that this technology develops.
81


CHAPTER VI
APPENDICES
A. Survey of the Profession
A statewide survey was conducted to determine how much experience LAs in Colorado have with geographic information systems, what their level of interest in them is, and what value they perceive them to be to the profession as a whole. A questionnaire was sent to 400 members of the American Society of Landscape Architects (ASLA) in Colorado. A brief article was published in the March, 1983 Newsletter, and the questionnaire was inserted. Of the 400 questionnaires sent, 106 were returned. A summary of the results is shown in Figure 22.
While 60% of the respondents used computers in their workplace, only 20% had ever used a GIS. Since micro-computers and minicomputers were by far the predominant computer system used in the^ workplace (77%), geographic information systems based on micros and minis are likely be the most feasibly purchased by LAs in the future. While the level of knowledge about GISs was low (only 8% understood most of their uses and capabilities for LAs), interest in further information was very high (85%). These results underscore the importance of this project and in further GIS literature and training geared toward the profession.
82


(CCASLA)
n= 100
Colorado Chipttr American 8oalaty ol Landaeapa Arohltaeta
loot
101
M
7c>Z
U>1
Sot
401
501
207.
IDT,
S*t
P';.V V :,v r-' $!
YES MO HIGH MOD. LOU1 HIGI+ MOD. MOME 1 YES MO
CSS ujokd fcoooa*F.
HI EHiiiNeevjta* v^ygjc-'
niTiq DCMCU MO«.lC f
U8C OF COMPUTERS IN THE INTERE8T IN FURTHER
WORKPLACE KNOWLEDGE ABOUT 018a EXPERIENCE WITH OlSa INFORMATION ABOUT QI8a
ro
00
SURVEY RESULTS
Figure 22


B. Glossary of Terms
Access The act of fetching an item from or storing an item in any computer memory device.
Accuracy The degree to which a measured value is known to approximate a given value.
Address An identification, represented by a group of symbols, that specifies a register or computer memory location.
Algorithm A finite set of instructions which, if followed, accomplish a particular task.
Alphanumeric A set of characters with letters, numbers, punctuation marks and special symbols.
Assembler A computer program that translates instructions written in a source language directly into machine language.
Auxiliary Memory Any computer memory or memories used to supplement main memory.
Batch Processing A method in which a number of data items or transactions are coded and collected into groups and processed sequentially. ^
Bit Acronym for Binary Digit, the smallest unit of information which can be stored in the computer.
Boundary General term for the division between two mapped areas.
Buffer The internal portion of a data processing system which serves as an intermediate storage between two different storage or data handling systems with different access times or formats.

Byte A group of adjunct bits that are operated on as one unit.
Card Image A representation in computer storage of the hole patterns of a punched card. The holes are represented by one binary digit and the spaces are represented by the other binary digit.
Cathode Ray Tube An electronic tube with a screen that is used in computer terminals to display input and output data. Also referred to as a CRT.
84


Cell The smallest region in a grid.
Central Processing Unit The central processing unit or CPE of the computer is that portion of the computer which is used to control the components of the hardware system.
Centroid The center point of a mass or polygon.
Chain A synonym for a string, e.g., "a chain of coordinates".
Character A letter, digit or other special symbol used for the representation of information.
Compiler A computer program that converts a source language into an object language.
Coordinate An ordered set of values that specify a location.
Core The most accessible information storage of a computer.
Cursor A movable part of an instrument that indicates (x, y) coordinates to the machine.
Data Base A set of data files organized in such a manner that retrieval and updating can be done on a selective basis and in an efficient manner.
Data Structure The arrangement and interrelationship of data.
Data Tablet A flat tablet which will output the digital position of a pointer placed at any position on its surface.
Digitization > The process of converting analogue or graphic data into digital form. Manual digitization involves the transformation of data by an operator with or without mechanical computer processor, while automatic digitization requires the use of an
automatic device, i.e., scanner, pattern recognition, character recognition.
Digitizer A device which converts maps into a digital format for computer input.
Direct Access Interactive systems employ direct or random access in which the access time is not related to the location of the data in the computer memory, i.e., data does not have to be serially or sequentially searched.
85


Editing The detection and correction of errors.
Encode The process of applying a set of unambiguous rules to transform data from its original form to some coded representation, usually digital.
Field A group of characters that is treated as a unit of data.
Fi le A variable number of records grouped together and treated as a main division of data.
Fixed Length Record Relates to a number in which various records must contain the same number of characters.
Format The specific arrangement of data in a record or f ile.
FORTRAN An acronym for FORmula TRANslation, a procedure-oriented computer programming language.
Geocoding The geographic coding of the location of data items.
Geographic Base File A coded network.
Geographic Coordinates A spherical coordinate system for defining the position of points on the earth.
Geo-referenc-ing * Planimetric coordinate system which identifies points on the surface of the earth. Systems include 1 atitude-1ongitude, universal transverse mercator, stable plane coordinate and land survey systems, etc.
Grid Coordinates Euclidean coordinate system in which points are described by perpendicular distances from an arbitrary origin, usually on an (x, y) axis.
Hardware % The physical components of a computer and its peripheral equipment.
Hard Copy Printed or paper copy of computer output. Commonly a paper copy of the information displayed on a computer video terminal.
Information Retrieva1 Methods and procedures used for storing and retrieving specific data and/or references based on the information content of documents.
86


Interactive Mode Allows users to directly interact with the information system to input and/or manipulate and retrieve information in a real time framework.
Interface The junction between components of a data processing system.
Intersection The region containing all of the points common to two or more regions or polygons. See also union.
Light Pen A device the size of a bail-point pen which is used for pointing to a location on a CRT screen. One of several types of interactive positioning devices including a mouse, joystick and tracking ball.
Line Printer An output device for computers which prints one line at a time. It can be used as a high speed listing or, by spacing symbols, as a plotting device.
Machine Language Instructions written in a code that can be understood by the computer without further translation.
Magnetic Disk A computer memory device on which data is available by random access.
Magnetic Tape A computer memory or storage device which will store a large amount of data, but th:ps data is only accessible in a sequential search.
Memory An organization of storage units (bits, bytes) retained primarily for information retrieval.
Minicomputer An inexpensive CPU with limited core capacity.
Natural Lan- A user-oriented language which can be used to
guage search the computer files by operators who have no programming experience.
â– \ Network A connected set of segments and nodes.
Node A point which is common to two or more segments.
Object Language A machine language that is output from a compiler.
Of f-1ine Processing is not directly under the control of the central processing unit.
87


On-line
On-line Processing is directly under the control of the central processing unit. All interactive systems operate on-line.
Optical Character Recognition The process by which printed characters are read by light sensitive devices for computer input. Also referred to OCR.
Overlay The superimposition of one map or digital image over another of the same area in order to determine data combinations or intersections and unions.
Periphial Input and output equipment used to transmit data to and receive data from a CPU.
Plotter An (x, y) mechanism controlled by a computer, generally for the recording of location or spatial information. Lines are drawn as a series of vectors.
Polygon Plan figure consisting of three or more vertices (points) connected by line segments or sides.
Program The implementation of a procedure by the use of a computer programming language. A program consists of a set of instructions which direct the CPU in the performance of a specific task.
Random Access The process of obtaining information on- data from a computer storage device where the time required for such access is independent of the location of the information most recently obtained.
Raster Scan A line by line sweep across a display surface to generate or record an image.
Real Time Processing which appears instantaneous to the person or the device controlling a computation.
Resolution N Measure of the ability of an imaging system to separate the images of closely adjacent objects. Also the smallest area at which data can accurately be identified.
Software The set of programs used to instruct the computer in problem solving; consists of the operating system programs and applications programs.
String A consecutive sequence of characters.
Terminal A device used to input or output data from a computer, often from remote sites.
88


Time Sharing The concurrent use of a computer system by more than one user or program by allocating short time intervals of processing to each active user. The response time is usually so fast, that each user is given the impression that the computer's resources are totally designated to his task.
Uniform Grid Square, rectangular, or hexagonal lattice grid coordinate system for recording geographic data.
Union The region containing all of the points in two or more regions or polygons. See also intersection.
Variable- Length Relates to a file in which the various records may contain a different number of characters.

89


FOOTNOTES
â– 'â– Julius ects?",
Gy. Fabos, "Paperless Landscape Architecture: Future Landscape Journal, 10, No. 1 (1983), 13-18.
2
"Proceedings of the Harvard Conference on Computer Graphics
anning and Design," Computer Graphics News, November 1983, Newsletter of the National Computer Graphics Association.
â– ^Gilbert H. Castle, "Regional and Site Specific GIS /sis", theme delivered at the National ASLA Conference, .napolis, Indiana, 1984. (unpublished paper).
^Geoffry Dutton, Ed., Harvard Papers; on Geographic rmat pon Systems , Geographic t pon Systems : Surveys ,
’iews, and Criteria, Vol. 2 (Cambridge: President and Fellows irvard College, 1978), p. 4.
c;
Comarc Design Systems, Wallace McHenry Roberts and Todd, lwest Colorado Council of Governments, "Draft Areawide Water ity Plan for Eagle, Grand, Jackson, Pitkin, and Summit lies Colorado" (Frisco, Colorado, 1978).
^Ibid.
Ibid.
8
Tom Elmore, Personal interview, 26 April 1984.
yIbid.
^Bruce E. MacDougall, Microcomputers in Landscape Ltecture (New York: Elsevier, 1983), p. 207.
■'■■'•Jack Dangermond, "Software Components Commonly Used in raphic Information Systems," working draft of a theme ented at various professional meetings during 1982. jblished paper).
^Ibid.
1 3
Julius Gy. Fabos, "Paperless Landscape Architecture: re Prospects?", Landscape Journa 1 , 10, No. 1 ( 1983), p. 16.
1 4
Brimer Sherman, Interactive Systems, Director of
eting, personal interview, 10 Oct. 1983.
90


Marble, D.F., ed. Volumes 1, Geographica1
Computer Software for Spatial Data Handling. 2, and 3. Ottawa, Ontario: International Union, 1980.
Monmonier, M. S. "Cartography, Mapping, and Geographic Information", Progress in Human Geography, 7(3), 420-428, 1983.
Monmonier, M. S. Computer-assisted Cartography: Principles and Prospects. Englewood Cliffs, N.J.: Prentice-Ha 11, Inc., 1982.
Moody, Howard, NWCCOG. Telephone interview. 1 May 1984.
Moore, P., ed. Harvard Library of Computer Graphics. Vols. 1-6, Laboratory for Computer Graphics and Spatial Analysis. Cambridge, MA: 1982.
Naisbitt, John. "The New American Contest: Thinking Globally and Acting Locally" (abstract).. £2L°I!t;Lers for the 80's: Response to Change. Washington, D.C.: American Society of Landscape Architects, 1980.
Novak, Michael. The Experience of Nothingness. New York: Harper & Row Publishers, 1970.
Phillips, Richard L. "Computer Graphics in Urban and Environmental Systems," Tutorial: Computer Graphics, Beatty, John C. and Booth, Kellogg S. IEEE Computer Society (Silver Springs, MD), 1982..
Puterski, Robert, Senior Cartographer, State of Colorado Department of Local Affairs. Personal interviews. 21 November, 6 December 1983, and 28 March 1984.
Robinson, Arthur. "His Master's Voice: Computers are Learning to Listen and Talk Back." Science 80, 1, No. 3. (1980).
Rowes, Ruch A. "Perception of Perspective Block Diagrams." The American Cartoqrapher, No. 5, No. 1 (1978).
Sherman, Brimmer, Director of Marketing, Interactive Systems. Personal interview. 10 October 1983.
Sleeper, David. "Technological Choices Can be Deceptive." Conservation Foundation Letter. Washington, D.C.: January, 1980.
Sonnon, David, Systems Analyst, Computer Data Systems, Inc. Personal interview. 17 November 1983.
95


Starling, Robert, Ph.D. "Future GIS Developments" Talk delivered at the Geographic Information System Workshop, Western Energy and Land Use Team, Ft. Collins, Co., 7 December 1983.
Starr, L., and K. Anderson. "Some Thoughts on Cartographic and Geographic Information Systems for the 1980s", Proc. Pecora VII Symposium. Falls Church: ASP, 1982.
Tomlinson, R. F., ed. Geographic Information Systems (Proc. UNESCO/IGU 2nd Symp. Geographic Information Systems,. Ottawa, Ont, Canada, Aug. 1972).
Tomlinson, R. F., and H. W. Calkins, and P. F. Marble. Computer Hand ling o.f Geogra phi ca 1 Data. Vol. 13, Na tura 1 Resources Research. Paris: The UNESCO Press, 1976.
U.S., Interior, Geological Survey, Overview and USGS Activities, by R. McEwen, H. Calkins, and B. Ramey, in R. McEwen, R. Witmer, and B. Ramsey, eds., USGS Digital Cartographic Data Standards, Circular 895-A, Reston, VA: USGS, 1983a.
U. S.
Interior, National Statement for the National Scenic Pennsylvania. Oct.
Park Service, Draft Environmental Impact River Management Plan. Upper Delaware and Recreational River/New York 1982.
Weber, W. "Geographic Information Systems -Reflections on the Future Development", Yearbook of Cartography, XIX, pp. 119-138, 19
A Review and International
79,
96


Full Text

PAGE 1

GEOGRAPHIC INFORMATION SYSTEMS: TOOLS OF EMERGING IMPORTANCE FOR LANDSCAPE ARCHITECTS AN INTRODUCTION TO THE TECHNOLOGY AND CRITERIA FOR EVALUATING AND IMPLEMENTING A GIS AT UCD COLLEGE OF. DESIGN AND PLANNING BY MCGOVERN A view of Mt. McKinley from the Northeast produced by the Comarc system.

PAGE 2

GEOGRAPHIC INFORMATION SYSTEMS: TOOLS OF EMERGING IMPORTANCE FOR LANDSCAPE ARCHITECTS An Introduction to the T echnology and Criteria for Evaluating and Implementing a GIS at UCD College of D esign and Planning by Ruth McGovern A thesis submitted in partial fulfillment of the requirements for the degree of Masters in Landscape Architecture The University o f Colorado at Denver May 1984

PAGE 3

THIS THESIS IS SUBMITTED AS PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR A MASTER OF LANDSCAPE ARCHITECTURE DEGREE AT THE UNIVERSITY OF COLORADO AT DENVER COLLEGE OF DESIGN AND PLANNING GRADUATE PROGRAM OF LANDSCAPE ARCHITECTURE ACCEPTED: (Committee Member•s Name & Title)

PAGE 4

CONTENTS LIST OF FIGURES . . I . INTRODUCTION A . Statement of the Proble m B . Purpose of the Project . C. Procedure for the Investigation. II. OVERVIEW OF GISs . A. GIS Components B . Overview of Representative Projects. l . A Hiking Trail Suitability Model 2 . The Northwest Colorado Council of Governments Water Quality Management Plan . . C . Information Flow . l. Data Gathering 2. Data Entry . 3 . Data Structure 4 . Editing and Updating 5 . Storage of Data on Tape or Disk. 6. Data Retrieval, Analysis, and Manipulation 7 . Output Formats and Output Devices. D . Classifications of Systems . l. Orientation of the System . 2 . Size of the System E . Considerations for LAs l. Advantages . . . l l 5 7 7 ll 12 14 14 18 28 29 31 34 37 39 40 4 3 46 47 4 9 50 50

PAGE 5

III. IV. v. VI. 2 . Disadvantages. 3 . Scale. . 4 . Accuracy 5 . Cost . 6 . Rethinking Our Old M ethods CRITERIA FOR EVALUATING GISs A. Establish Your Needs and Determine Your Requirements. B . Determine Appropriate Size and Orientation of System. . .. l . Size of the System 2 . Orientation of the Syst em. C . Evaluate Reliability of the Vendor D . Appraise the GIS for Ease of Use and Flexibility. E . Cost Considerations .. RECOMMENDATIONS TO UCD LA PROGRAM. . . A . Needs/Requirements of the UCD LA Program B . Appropriate System Size and Orientation for UCD. l . Orientation of the Syst em . 2 . Size of the System 3 . Cooperation with Public and Private Agencies . 4. Incorporation Into the Curriculum. SUMMARY AND CONCLUSIONS . APPENDICES A. Surve y of the Profession 52 54 55 56 57 59 59 61 62 63 64 65 67 68 68 70 70 71 74 75 78 82 82

PAGE 6

B . Glossary of Terms .. FOOTNOTES . . BIBLIOGRAPHY 84 90 93

PAGE 7

Figure l 2 3 4 5 6 7 8 9 10 ll 12 13 14 15 16 17 18 1 9 20 2 1 22 FIGURES GIS Applications Process. GIS Hardware Components. Overlay. Hiking Trail Suitability Mod e l NWCCOG R egion 1 2 . NWCCOG Project Structure NWCCOG Study Scale Conversion. NWCCOG Study Ov erlays to Define Landscape Units .. NWCCOG Study Flow of Mappe d Data Information Flow . Data Relationships Data Entry T echniques. Type s of Spatial Data. Gridded Data Slivering. Interactive Editing F eature s M easure m ent Functions. Digital T e rrain Analysis Functio n s GIS Outpu t . Criteria for Evalu a ting a GIS. Survey R esults . l Page 3 8 1 3 15 17 19 21 19 25 25 30 . 32 . 36 36 . 36 . . . 38 38 41 41 44 6 0 83

PAGE 8

CHAPTER I INTRODUCTION As landscape architects, we vlew our rol e ln society as one of great responsibilit y . We charge ourselves with "stewardship of the land" (Fein, 1972), "design and management of the landscape (Killpack, 1980), and "dedication to the land and its people" (Marshall, 1973). This responsibility requires a thorough understanding of our biophysical and socio-cultural systems, as well as an understanding of the relationships and interactions of these systems. Getting an overall view of these relationships is no small or easy task, but is nevertheless one of the most important steps in our design and analysi,s process. The initial process of gathering, organizing, a nalyzing, and comparing natural and cultural information has been estimated to take as much as two-thirds of the total time involved in a project (Killpack, 1980). As we have more and more information to handle, it becomes increasingly important to improve our accuracy and efficiency. If we are to allow more time for thinking and creating, w e must look to new methods of doing things. While the increased complexity of our world creates problem s , n e w opportunitie s are arising, ln the words of Julius G y . Fabos, to "restructure the form and mode of LA practice" ( 19 8 3). Fabos suggests that landscape arc hi tectu re is one of the

PAGE 9

few remaining professions to escape the computer revolution. "Perhaps over ninety percent of our routin e planning, design and management activities are still done in the way ln which they have been done for decades.", h e says . l The computer is a n e w tool which has the potential to greatly aid and expan d the field of landscape architecture. One compute r r esource which has the capacity to be of tre m endous help to the landscape architect is the geographic information system (GIS). A GIS is a computer system which is designed t o store, process, analyze, and generate graphical output of information that is tied to a particular geographic location. In a GIS, locational data such as latitude and longitude are manipulate d along with attribute data such as vegetation and soil types. Som e special f eatures for landscape architects are map overlays and composites, of special areas and distances, and creations of buffer zones around specified areas . With the aid o f a GIS, for exampl e , an LA can build and display map models for site selection, or models for projection of development and change in an area. In addition, a GIS can perform such functions as delineations of view sheds, c a l c u l at ions of w a t e r f l ow and run-of f , and c a l c u l at ions of density. Som e other use s of GISs are liste d in Figure l . A GIS, then, can greatly r educe the time and drudgery involved ln compiling this kind of data, increase the l e v e l of detail, g e n erate o u tput s ln forms which are attractive a n d eas y t o compre h end, and allow the landscape architect to focus the majority of his/her efforts toward d esign considerations. 2

PAGE 10

SITE PLANNING URBAN AND REGIONAL LAND USE PLANNING LAND USE INVENTORIES MASTER PLANNING SOCIOECONOMIC STUDIES NATURAL RESOURCE PLANNING REGIONAL ENERGY -IMPACT AND ANALYSIS LOCAL EIS ANALYSIS PREDICTIVE MAPPING -PROJECTING POPULATIONS ESTIMATING RECREATION ACTIVITY PREDICTING STORM RUN OFF PREDICTING EROSION HAZARDS ROUTE SELECTION FOR HIGHWAYS ROUTE SELECTION FOR BUSES AND OTHER PUBLIC TRANSPORT.A TION TOPOGRAPHY ANALYSIS FOR DEVELOPMENT STATISTICAL MAPPING (CENSUS, CRIMES, ACCIDENTS, ETC.) UTILITY FACILITY MAPPING AND MANAGEMENT GIS APPLICATIONS Figure 1 3

PAGE 11

In spite of their apparently enormous potential for LAs, geographic information systems have been slow to enter the field of landscape architecture. At the recent Harvard Conference on Computer Graphics in Planning and Design, Chairperson Madson noted, "Geographic information systems have been widely developed and used in the fields of geography and cartography for more than a decade, but have been used very little ln the fields of architecture and landscape architecture. The time and technology, however, are now ripe for the latter two fields."2 Until recently GISs have been out of reach for all but institutions and agencies with access to large computers and the funding to afford the necessary data manipulations. Advances in both computer hardware and software have turned this situation around. Today, many useful GIS applications can be performed relatively cheaply using mid-sized (mini) and even small, desk-top (micro) computers. In addition, the conventional, time consuming and costly methods of formatting and inputting raw data into the computer (e. g., digitizing) are gradually being replaced by automated methods of data input such as scannlng. Finally, GISs are becoming incre_asingly user-friendly and thus understandable and usable even for the computer novice. Geographic information systems, then, are emerging as viabl e tools for landscape architects they have the capacity to provide an abundance of varied natural and cultural data from which we, as r esponsible managers and designers of the landscape, 4

PAGE 12

should draw in making our design decisions. By taking much of the drudgery out of the process of assimilating this kind of information, as w ell as by s ignif icantl y improving accuracy and efficiency, GISs can provide LA' s with more time for thinking and creating; for generating and evaluating d esign alternatives. A . Problem Statement Although computer technology has advanced to the poin t where geographic information systems are affordable and have the potential to be tremendousl y valuabl e tools for landscape architects, GIS technology still remains large l y inacccessible to m embers of the profession. There appear to b e two primary reasons for this: l) Virtually no information is currently available 1n the GIS literature which is specifically directe d toward LAs. Learning about GISs and precisely what they can do for LAs ofte n involves struggling through much jargon-riddled technical writing. 2 ) There are few outlets currently available to LAs for learning about and gaining hands-on experience with GISs. The second problem of allowing LAs direct exposure to GISs is now b eing addressed at the university level, but so far only at a handful of universities on the east and west coasts. Here in Colorado, the need for LAs to make use of GIS technology is particularly strong given the problems w e face with t h e rapid population growth along the front range . My survey of Colorado 5

PAGE 13

professionals (see Appendix A), found that the current level of knowledge about GISs is quite low, but that interest in learning about GISs is high. Another significant finding was that virtually all LA firms either own a micro-computer or have plans to purchase one in the near future . Given these facts, it is apparent that the use of GIS technology by LAs is likely to surge 1n the near future if given the proper impetus. The LA program at the University of Colorado at Denver is in the unique position to spearhead this imminent surge because .l) it is the only accredited MLA program in Colorado, and 2) the Denver metropolitan area contains the state' s greatest concentration of LAs. By procuring all the necessary components of a user-friendly and versatile GIS, the UCD LA program would be able to offer students and area professionals a valuable outlet for learning how to make use of this rapidly emerging technology. This would likely have several beneficial effects: l) It could spark increased utilization of GIS technology by LAs in the area. 2) It would raise the general level of computer literacy for MLA students --an issue of some importance to the degree program accreditation committee. 3) It would promote better linkage with those students and professionals in fields with whom LAs are associated who are making increasing use of computer technology . 4) It would anticipate the inevitable burgeoning of 6

PAGE 14

interest ln GISs among LAs as the tools become more widespread. The system would already be in place in preparation for meeting increasing demands for GIS training. 5) It could serve as a resource for area professionals ln much the same manne r as a research library. 6) It would be a general asse t to the university and could enhance the LA division' s image. B. Purpose of the Project It is apparent that the time is now ripe for implementing at UCD just such a resource as I have suggested here: an easy-to-use and versatile geographic information system. The purpose of this paper then, is to help pave the way for this implementation b y providing l) an overview of GISs, written in terms to those having no prior computer or cartographic experience, emphasizing the characteristics and practical applications of GISs as they relate to the landscape architecture profession, 2) a set of criteria and a format for evaluating various GISs based on specific needs, and 3) recommendations for implementing GIS t echnology at the UCD College of D esign and Planning. C . Procedure for the Investigation Th e organization of this project lS diagrammed in Figure 2 . In pre s enting an ove r v i e w of GISs, I will draw from several different sources --from GIS literature geared toward computer specialists, geographers, and cartographers; from intervie ws with 7

PAGE 15

DEFINITION PROJECT REVIEW INFORMATION FLOW CLASSIFICATIONS 9F SYSTEMS OVERVIEW OF GEOGRAPHIC INFORMATION SYSTEMS CONSIDERATIONS FOR L -ANDSCAPE ARCHITECTS RECOMMENDATIONS TO U -CD L-AN OS CAPE ARCHITECTURE DEPARTMENT PROCESS Figure 2 8

PAGE 16

varlous professionals knowledgeabl e about the syste ms , including GIS t echnicians, users, and vendors; and, finally, from my own p ersonal experience in working with a GIS at WELUT (Western Energy and Land Us e T e am , a division of the U . S . Fish and Wildl ife Service in Ft. Collins, Colorado ) . Th e overview section is structured from the gen eral to the specific. The first portion o f the overvie w illustrate s the relationship of GIS t echnology to anothe r comput e r graphic technology and defines the fundamental components of a GIS. The n ext portion reviews GIS projects representative o f projects use d by landscape architects, discusses the projects' formats and purposes, a n d explores the types of analyses used. The third portion of the overview goes more specifically into how a GIS works, exploring its functions and capabilities. This is designed both to provide a selfcontained introduction to GIS technology for interested LAs and to serve as a foundation for those individuals at UCD who ma y wish to make use of the ensuing section on GIS evaluation criteria. The set of criteria in the third chapter of this report is intende d to provide a framework for evaluating the suitability o f GISs for LA applications. This is intended to provide a process for interested LAs t o follow in evaluating and s e lecting a GIS. This framework i s then utilized in this report as the basis for establishing goals and selecting a GIS for the LA program. The focus is, of course , on the LA D epartment, but as much as possible the goals and directions of other departments in the 9

PAGE 17

college are considered as well. Give n the complexity of the decision-making process in determining an appropriate system, specific hardware and software recommendations are beyond the scope of this paper. Rather, the intention is to explicate the essential considerations and to explore various options available when choosing a GI S . It is hoped that this will provide the UCD Landscape Architecture D epartment with direction for imple m enting geographic information systems technology into the program, som e initial impetus, guidelines for evaluating systems, and recommendations of issues for further consideration. 10

PAGE 18

CHAPTER II OVERVIEW OF GISs To show how geographic information systems can b e most useful to landscape architects, this section looks at how they work, how they are used, and examines some special features and problems r elate d to the LA profession . Section A presents the components of a GIS. Section B looks at some actual GIS projects; examines their purposes, how they were structured, and how they were used. Section C describes in detail the flow of information in a GIS, including data input, storage, retrieval, analysis, and output. Section D discusses the requirements and special features of GISs as they relate to size type of system. Finally, E sums up some considerations about GISs specific to the LA profession. Many of the functions o f a GIS overlap with those of a computer aided design (CAD) system. For example, both have data base management systems and extensive graphics modeling capabilities. The'difference between a CAD system and a GIS is that the focus of the CAD system is on design per se, whereas the focus of a GIS lS on analysis and interpretation of geographically based data. Th e GIS is neverthe less integral to the design process in that it provide s a context in which d esign considerations should take p lace. GIS technology combines mapplng and analysis capabilities so l l

PAGE 19

that land use professionals can interact directly with geographic information and quickly obtain accurate answers to pertinent "what if" and "where" type questions. 3 The questions might r elate to such things as the best uses of a particular site , a comprehensive land use plan, or a coastal zon e management program. GIS analysis may encompass mapped information, r e lated tabular data, t ext information, and even satellite data. A . GIS Components A GIS may be a dedicated, stand-alone system, may be part of a network of computers used for GIS purpose s , or may be only one o f several functions performe d on a computer. The principle hardware components of a GIS are shown in Figure 3. These include : 1) a computer, 2 ) data storage device such as a disk, 3) data transfer device such as a tape drive, 4) digitizer for e ncoding maps and aerial photographs, 5 ) graphics terminal (monochromatic or color), for data display and i n teractive communication, 6) line printer, used to print tabular data and generate low resolution graphics, 7) plotte r for generating graphics "hard copies" . GIS softwar e refers to the programs that perform the work r equest e d b y the user. The software contains 1) extensive data base management system s to handle data input, search, and 1 2

PAGE 20

dele etorege end ...• I I 1 I I I I I I I I computer GIS HARDWARE COMPONENTS F lgure 3 pi ott • r

PAGE 21

retrieval, and 2 ) graphics capabilities to format data for output in the form of maps, charts, cross sections, and 3-D images. B . Overview of R epresentative GIS Projects This section reviews two diverse GIS projects, examines how the projects were designed and structured, and looks at some of the specific GIS functions that we r e used. l . Trail Suitability Model The first project is a hypothetical one; it illustrates the use of a GIS to build a hiking trail suitability model. The final product from this project would b e one map of an area delineating areas of high, medium, and low suitability based on a set of criteria established by the designer. It is used here because of its relative simplicity a n d because it illustrate s a GIS function often used manually by LAs to build suitability models the overlay. The purpose of an overlay is to compare two or more maps or the mes in a given area. For example, Figure 4 shows the result of overlaying a zoning map on a flood plain map. The GIS would creat e a n e w map showing exactly which portions of each zoning category we r e i n each portion of each flood category. With the overlay function, maps can be produced which show l) areas ln common between two or more maps or data sets, 2) areas not ln common between these maps or data sets, and 3 ) areas both ln common and not ln common on two or more data sets. 4 Such questions as how much national forest land lies i n a given county 14

PAGE 22

OVERLAY Figure 4 15

PAGE 23

or how much cleare d land lie s 1n a propose d d evelopment are simple examples that require an overlay manipulation. Figure 5 diagrams the process that could be used to build a hiking trail suitability model. This is a standard process employe d in setting up a strategy for building a suitability map from a set of overlays. The firs t step is to d efine what the fina l map will b e . The s econd ste p 1s to identify the components o f the fina l map. In this case , erosion potential, hiking difficulty , a esthetic quality, and cost were identified as e l e m ents which would d e t ermine the ultimate suitability for a hiking trail. ( E 1 ernents can b e adde d or d e l eted depending on what the designer and client d ecide is important.) The third step is to identify a s e t of sub-components. For example, v iews, noise, and surface texture were as sub-components which comprise the component aesthetic quality. Fourth, additional sets of sub-components are again identifie d for each new category until one arrives at a data type or substantive map is. Determined as affecting v i ews in this step, for example , were elevation, vegetation, and cultural features each o f which is available as a quantitative map. To d e t ermine hiking trail suitability in an area, s o m e s u b jective d ecisions need to b e made. The firs t subjective decisi o n i s defining what components det ermine suitability . T h e second subjective decision i s in ranking the four components said to determine suitability (i.e., whic h is the least /mos t importan t factor). To s e t up a ranking structure, w e i ghts or values are 16

PAGE 24

HIKING TRAIL SUITABILITY ENVIRONMENTAL SENSITIVITY HIKING DIFFICULTY AESTHETIC QUALITY COST VIEWS SURFACE TEXTURE HIKING TRAIL SUIT ABILITY MODEL Figure 5

PAGE 25

given to each of the components at each step. The final outcome , then, will be a map with areas weighted as to their hiking trail suitability and with each area having a quantitative value. The model is executed in reverse order (i.e., right to left on the diagram), so that, in this case , elevation, vegetation, and cultural features are overlaid to make up the new ma p "views", which is overlaid with other maps to create the aesthe-tic quality map, which again is a sub-map of the final hiking trail suitability map. 2. The Northwest Colorado Council of Governments Water Quality Management Plan The second GIS project reviewed here .was an actual project involving a s1x county 9,064 square mile -area in Colorado Region 12 (Eagle, Grand, Jackson, Pitkin, Routt, and Summit Counties). See Figure 6. _ , In 1976, Northwest Colorado Council of Governments (NWCCOG) contracted a GIS company, COMARC Designs Systems, to assist in preparing a Water Quality Management Plan for the six county region. This GIS project is reviewed here because it l) was a Colorado project, 2) is repre s entative of the potential scale and scope of GIS projects, 3) is a completed project, with enough lapsed time to evaluate its overall effectiveness, 4) exemplifies the complexity and significance of initial project p lanning and structure , and 6) was subcontracted by landscape architect Ian McHarg. The project is first discussed in terms of the original 18

PAGE 26

NWCCOG REGION 1 2 Figure 6 Three Lakes land Use NWCCOG STUDY SCALE CONVERSION Figure 8 l 9

PAGE 27

goals, structure and final implementation. A summary reviews the project's strengths and weaknesses based on its original intent. The objective of the project was to devise a Water Management Plan to comply with Section 208 of the Clean Water Act. The plan was also intended to develop definitive, site specific water quality management policies for the entire area, to accommodate new legislation and guidelines, improved mitigation measures, and more refined data.5 The region's water resources are of national importance. It contains the headwaters of the Colorado River and headwaters feeding into the North Platte. Millions of people in several states and Mexico are dependent on these for their water supply. Within the area are high mountain meadows, exposed bedrock summits, deep canyons, forested mountain slopes, valley meadows, cliffs, and broad intermontane basins. It e hcompasses scenic land, a national park, ski resorts, and deposits of valuable mineral resources. Demands for additional recreation, housing, and mining, coupled with a sensitive ecosystem, create a tremendous potential for conflict. Because of the size of the project, the amount of data, and t h e number of agencies involved, (the National Park Service ,the National Forest Service, private landowners, and county planners}, the overall project structure was complex. As shown in Figure 7, the main components were l} inventory of data sources, 2} program definition, 3} data collection/encoding and formatting, 4} data analysis, 5} pol icy 20

PAGE 28

INVENTORY OF DATA SOURCES II PROGRAM DEFINITION II DATA COLLECTION ENCODING AND FORMATTING DATA ANALYSIS. I POLICY FORMULATION I COMPREHENSIVE PLANNING IMPLEMENTATION NWCCOG PROJECT STRUCTURE Figur e 7 21

PAGE 29

formulation, 6) comprehensive planning and 7) implementation.6 data sources. The first step was a comprehensive inventory of data sources. An environmental atlas was prepared, some 3000 sources of information on the region. Program definition. Program definition was based on public input, policy requirements, review of potential activity areas, and design of scme prctotype projects. It was recognized at the outset that water quality management must be viewed as a part of the overall planning effort. It was therefore to consider other planning issues and the extent to which they would impact or be impacted by the Water Quality Management Plan. A comprehensive rlan for the six county region, then, would be developed which would also look at development and growth for specific study areas within the region. Data and From the inventory, and review of the data needed tc prcduce a regional water quality management plan, a data base was designed using new and existing data. U.S. Geologic Survey wc: E contracted,, for example, to perform a water quality inventory o f surface water cf the while data en soils information were gathered frcm existing Soil Conservation Survey All data were digitized and stored in the computer lsing CGDarc's Geographic Information SystEm, COMPIS. entire 9,064 square mile area included:7 22 Deta encoded for the

PAGE 30

for this was the overlay method described in the previous project review, the hiking trail suitability study. All the inventory maps were overlaid to determine all combinations of data and their frequency of occurrence. This was carried out in stages to allow for intermittent interpretation and analysis by the planning team. Figure 9 indicates this initial process of determining landscape units. First, the sub-basin maps were computer overlaid on the data base to break it down into nine separate sub-basins. Second, within each sub-basin, geologic formations were overlaid on the headwaters map to determine the occurrence of groups of geologic types in headwaters and nonheadwater areas; these were characterized as "Level One Landscape Unit". Third, these "Level One Landscape Units" were compared by the GIS to other maps in the data base to determiQB all the combinations of characterizations and extent to which each existed in the region. Finally, "Level Two Landscape Units" were defined. These represent the ultimate division of the area into ecological units -defining areas distinct from one another in their hydrologic response to change. ' Policy formulation. Figure 10 indicates the flow of mapped data through the computer. Each of the landscape units was evaluated in light of the regional water quality policies and was used as the basis for how policies should be applied to specific development sites. Comprehensive planning. Public and advisory board meetings )4

PAGE 31

Topography Geologic Formations Climate Land Use/Cover Type Potential Activities Zone (PAZ) Soil Associations Surface Water Headwater Areas Sub-Basins Reservoir Catchment Areas Source maps had a variety of scales ranging from 1:1,000 to 1:250,000. Once the maps were in the system, coordinate conversions and scale changes were made by the GIS as necessary. Maps were merged with other maps and displayed with all the data at a consistent scale. From topographic data, the system computed slope, aspect, and elevation. In addition to the regional data base, more detailed data bases were established for four specific study areas within the region: The Aspen-Snowmass Area in Pitkin County, the Lower Blue River Area in Summit County, the Three Lakes Area in Grand County, and the Walden-Coalmont Area of Jackson County. It was thought that these four areas faced most and that they contained specific problems represented throughout the region. Therefore, their data bases were more comprehensive than the regional data base. A particularly valuable application of the computerized data base was extracting information for any locale by zooming into a specified area. For example, land use was encoded for the whole region, but could be retrieved and displayed at a smaller scale. Figure 8 on page 19 shows land use being displayed for one county and then for one USGS 7-l/2' quadrangle within the county. The initial analysis involved determining what ecological units occur in the region. The process employe d 23

PAGE 32

NWCCOG STUDY OVERLAYS TO DEFINE LANDSCAPE UNITS ... . ___ Figure 9 --, lull.wtlly Growl, NWCCOG STUDY FLOW OF MAPPED DATA Figure 10 / Oro wth / W•a.r 0\.laiMy I lfil&erlKe IJ") N

PAGE 33

were held at nearly every stage in this project's process. In preparing the Comprehensive Plan. in particular, the public became involved. A series of workshops were conducted in each county where the public was given the opportunity to comment on the proposed policies and then view those policies within the overall context o f planning for their county. During these workshops, participants established criteria to be used for producing computerized development suitability maps. Using the Geographic Information System, they established population projections relating to specific growth inducements and locations. They developed phased growth projections for their county on the computer, resulting in a map of future growth in the region. Additionally, these maps were overlaid by the GIS onto the Level Two Landscape Units to determine potential conflicts between these growth patterns and each of the proposed quality management policies. A report was then prepared for each county documenting these conflicts and listing mitigation measures which could be used. The original idea 1n utilizing a GIS for this project was to develop a regional comprehensive plan with ' several site specific development guidelines and to maintain a regional data base which could be used in subsequent years for other regional planning issues. It was thought that the regional data base could be used by different agencies, edited and changed through the years, and result in more consistency between various planning efforts. 26

PAGE 34

While some of these goals have been partially or fully realized, a lack of foresight at key points initially caused problems in making use of the more site specific data. Housing development is in great demand in the area and was to be studied using the GIS data. A large percentage of the region is government owned (primarily U.S. Forest Service) land, so there are limited areas where development can occur. Another major limitation involves the installation of septic tanks. Soil constraints in certain areas prohibit development because of severe leaching. Although soil data were a part of the data base, they were on too gross a scale (standard Soil Conservation Survey maps) to apply to the five-acre lot size being reviewed in several county planning processes. In addition, soil profile and depth information was not included, rendering the data totally useless to the specific analysis needed for the sites' , development review standards. Thus although the system could zoom into smaller areas for purposes of site specific analyses, the base data was too coarse to allow credible use.8 In the seven years since the comprehensive plan was devised, it has been used successfully on a county-by-county basis. Several couhty planners involved in the original study noted that the GIS was particularly useful for public meetings -both in promoting discussions via computerized future projection scenarios, and in credibility and clear documentation of the analysis process.9 The original data base has been utilized and appended by outside agencies. A region-wide traffic study used and added to 27

PAGE 35

various aspects of the data base. In addition, the data base was used by a utility company to determine the most suitable routes for transmission lines in the region. This project exemplifies the potential scale and scope of GIS projects. The project used the GIS for large scale analyses, for promoting public participation, and inter-agency cooperation. It illustrates the use of various manipulative techniques typical in GIS analysis. Most notably, those include various topographic terrain analysis (slope, aspect, elevation), overlay and composite analysis, future projection analysis employing the "what if" query, map scale conversions, and locale "zooming" for analysis at smaller scales. The project also demonstrates serious problems that can arise from a lack of forethought, underscoring the impprtance of . , initial project planning, of thinking through what the data will be used for and determining the level of depth and detail that will be required. C. Information Flow There are five distinct steps in the flow of information when using a GIS. These are: 1) data gathering, 2) entry of raw data into the computer (e.g., digitizing), 3) storage of digital data onto magnetic tape or disk, 4) selective data retrieval and an a 1 y s i s , a n d 5 ) outpu t of pro c e s s e d data. Ed i tin g o f data can take place between steps 2 and 3 and between steps 3 and 4 . Often much of the pertinent data for a project will be available 28

PAGE 36

already in digital f orm, in which case the time consuming efforts involved in step 2 can be reduced or eliminated. Figure ll shows these basic steps in the GIS flow o f information. Each of the steps is examined in detail below. l. Data Gathering Data for g eographic information systems can be gathered from a variety o f sources: maps, aerial photographs, field surveys, assessor' s records, census tapes, as well as from any standard source used by LAs in their analysis process. Since the data from standard sources need to be converted into machine readable (digital) form, (see Data Entry, below) the trend is toward using data already available in digital form. In the U.S., the United States Geological Survey (USGS) now distributes topographic and land use information in digital form, the Soil Service digitizes their maps, and LANDSAT (satellite derived data) is available on digital tapes.10 Many state agencies have converted their natural resource information to digital form as well. The amount and type of data used in a GIS, of course, is dependent on the scale and scope of the project and is determined ' by the users. The spatial data of a GIS is most commonly divided into two basic categories: l) locational data, indicating the size, location, and shape of a unit of data; and 2) non-locational o r attribute data, giving the characteristics (soil type, land ownership, tax rates, etc. ) for each unit. A third characteristic being used more and more is that of time 29

PAGE 37

EDIT AND/OR UPDATE DATA 1. GATHER DATA 2. 3. '/ ENTER OAT A INTO THE COMPUTER (DIGITIZE) _, STORE DATA ON TAPE OR DISK ' / RETRIEVE SPECIFIED 4. DATA FROM STORAGE ,I/ 5. MANIPULATE DATA 6. OUTPUT PROCESSED DATA INFORMATION FLOW Figure 1 1 3 0

PAGE 38

(Dangermond, 1982). Figure 12 is a conceptual diagram of there-lationship of these three elements, i . e . , locational data, at-tribute (non-locational) data, time. Since locational and at-tribute data often change independently of one another with respect to time, management of spatial data is most effective when locational and attribute data are managed independently. 2 . Data Entry For standard da:ta such as maps and photos, it is necessary to convert the spatial data into machine readable form before it can be entered and placed in storage in the computer. There are a number of techniques used for this purpose, including: a) manual digitizing, b) automatic entry devices, c) automatic line following techniques, d) optical scanners, e) manual cell encoding, f) survey document input. These are described below and explained diagramatically in Figure 13 , page 36 . a) Manual digitizing is the most commonly used data ' conversio n method today. The digitizer is an electromagnetic, electrostatic or electromechanical device which converts maps or other data into digital format. The digitizer consists of a flat table with an internal (x, y) matrix and a cursor. (See page 13, Figure 3 . ) The operator traces the cursor, or "point locator", over the map and records key points, lines, and areas which are 31

PAGE 39

1 GEOGRAPHIC DATA LOCA TIONAL DATA NON-LOCATIONAL DATA + DATA RELATIONSHIPS Figure 1 2 32

PAGE 40

read directly into the computer. This process is quite time consuming, usually requiring a greater investment of time than all the other aspects of a GIS project combined. b) Automatic entry devices refer to a variety of emerging technologies and instruments which directly capture and record spatial data automatically on site (referred to as "real time" data collection). These include satellites such as LANDSAT and related image processing technologies, digital photogrammetric systems, airborn geophysical equipment and "auto surveying" devices which capture coordinates as they pass over terrain.12 The main advantage of these systems is the timeliness and accuracy of the data. However, at this time, most of these systems work with a relatively large scale (e.g.,l.l acre resolution for LANDSAT ' s first three satellites), making them impractical for certain applications. This problem is being remedied, though, as LANDSAT's most recent satellite (NASA Module 8-21980) is capable of producing spatial resolution of 30 x 30 meters.13 c) Automatic line followers (ALF) are laser driven devices which are used to scan off a continuous string of coordinates ' associated with lines of a map. The ALF traces the lines by sensing the difference of light reflectance between the line and the background. The devices have the big advantage of time and manpower savings, but currently are uneconomical for most applications. d) Scanning devices, similar to ALP 's, automatically 33

PAGE 41

convert printed material into a machine readable matrix form by sensing different intensities of reflected light. They also hold promise for the future, but are still in the initial stages of practical usefulness. e) Manual cell encoding involves encoding points in a grid assigning dominant features a higher numerical value. These are wr i ,tten on a coding form, then keypunched on cards and fed in to the computer. Used in the early stages of GIS development, this process is time-consuming and impractical today. f) Input of survey documents is typically done through the manual key entry of x, y reference points, bearings, and distances relative to specific features desired for encoding. Features commonly encoded include ownership boundaries, roads, and buildings. These are especially useful when working with land ownership, census data, grid utility mapping, but.are little used for LA analysis. Emerging automated data entry technologies and positively alter the data conversion process. are that it is only a matter of five years technologies are viable.14 ' 3. Data Structure will radically Speculations before these Data structure influences the type and volume of data, the way in which they are collected, their resolution (distinction between closely related objects), validity, and ease o f manipulation. The three types of spatial or locational data used 1n a GIS 34

PAGE 42

are points, lines, and areas (see Figure 14). Points are used to identify locations of features which are so small as to have no areal context (waterfalls, wells, mines, etc.). Lines are used for linear data such as streams and roads. Areas, or.enclosed spaces, are the most common geographic data and include land use types, vegetation types, soils, counties, etc. The two types of areal data structures are grid and polygon. In a grid system, (also referred to as a raster system) a uniform grid, typically square, 1s super-imposed over the study area, and the predominate attribute of each grid cell is entered into the system (see Figure 15). This allows certain types of manipulations which can' t be done in polygon format (i.e., the matrix manipulations used with most programming languages), or which can be done more quickly in a grid format. The major disadvantage is that, because of the cells, exact locations may be lost. This kind of process is best done for very large areas where some kind of generalization is necessary or desirable, or 1n situations where computer memory is limited. In a polygon system, the attributes are directly and fully The two types of polygon systems are polygon loops and arcs.l5 Polygon loop systems represent the first generation o f polygon systems. Each polygon is digitized separately and stored as an individual record. Consequently, the boundarie s of adjacent polygons are digitized twice. Since it is almost impossible to exactly duplicate the line each time it is 35

PAGE 43

Q 1111 t..C.tMtl'-' • . ' llt&Wit .. lll DATA ENTRY TECHNIQUES Figure 13 GRIDDED DATA Figure 15 I I I I I . Original Grid Over lay Grid Cell Xap 36 . -LINE [;:J POLYGON TYPES OF SPATIAL DATA Figure 14

PAGE 44

digitized, an undesirable phenomenon known as slivering occurs (see Figure 16). This causes overlapping lines or unnecessary spaces between polygons. With the arc structure, only the boundaries between polygons are d i g i t i z e d , and t h e 1 e f t and r i g h t po 1 y go n 1 abe 1 s ( i . e . , attribute names) are entered at the digitizing stage. Since each line is digitized only once, slivering is not a problem. For this reason, the arc structure has become the more desirable design . . Also important in the structure of a geographic information system are the location identifiers, or gee-referencing systems. Latitude and longitude and other x , y coordinate systems are the m ost c o mmon methods used for indicating spatia 1 1 oca tion. Geographic data can also be represented entirely without coordinates, using principles of graph theory topo-graphical relationships to express relative location. This type of gee-referencing system is most often used for concepts relating to networks (streets, rivers, etc. ) or when any graphing function is required. Some GISs incorporate both of these geo-referencing systems. ' 4 . Editing and Updating Both while the data is being converted and entered in the computer, and after this digitizing process is complete, the data needs to be edited and checked. Some checks are done by the software, making this process automatic. Those syste ms which do offer this feature are obviously more convenient and can save a 37

PAGE 45

SLIVERING Figure 16 Figure 1 7 38

PAGE 46

tremendous amount of time. Some common elements of initial checking process are: a) plotting or printing out digitized data for visual editing; including checking it with the original for discrepancies, b) checking for polygon closure, c) checking with surrounding data to make sure they match (referred to as edge match), d) deleting duplicate or unnecessary lines (slivers) or points, e) checking polygon labels. The updating frequency depends somewhat on the data type. Certain types of data, such as geologic features, are fairly stable, and if entered accurately needn't be changed for years. Other data, such as vegetation or demographic da t)3., , change frequently. Editing features vary from system to system. Some basic interactive editing features to look for are listed in Figure 17. 5. Storage of Data on Tape or Disk As data are entered into the computer through means of ' manual or automatic digitizing techniques, the Data Base Management System (DBMS) of the GIS software takes care of storing the data on magnetic tape or disk. In the process, it sorts the data and files it in such a way as to allow for the greatest efficiency when later searching for specific data types. If the user is fortunate enough to have located a source for 39

PAGE 47

pertinent data which is already in digital form, this is the stage at which the user can transfer the data to his/her system, re-formatting the data as necessary. 6. Data Retrieval, Analysis, and Manipulation Computer analysis is, of course, the primary function of a GIS. This section describes the various analytic and data processing functions that can be performed by a GIS. In terms o f minimum data function, there is no standard definition of a GIS. Tomlinson suggests that if the manipulations of search, measurement, and comparison are not needed, then a GIS is not needed.1 6 Most GIS functions include combinations of these three components. Searching is the capability to read an entire file and retrieve any specified set of data by location, attribute, or specified attribute value. Measurement functions are shown in Figure 18. These include calculations of area, perimeter, distance, volume, and direction. Comparison is probably the most. powerful capability of a GIS for LAs. It uses descriptive or location data to determine relationships based on criteria from one.or both types of data. ' The most common comparative techniques are overlaying and compositing. (The technique of overlaying, incidentally, was pioneered by well-known landscape architect Ian McHarg.) Overlays are very useful in building suitability maps. Relationships of maps or attributes can be described quantitatively and comprehensibly in a matter of minutes. 40

PAGE 48

' y + I + t-)' t+ v + tTo cal Number (Straight) (Aren) Cross Section &?:/ y /r-. rV y /y4 • • 1 1 'u( ce I fb .... I Pol. y 1 1 ,, "1-0 5;u.-f .. ca • 'Z Poinc-io-Polygon (Curved) 'Perimeter) Area I'OINTS DISTANCE AREAS VOLUMES MEASUREMENT FUNCTIONS Figure 18 --+--+--+'10 '171'" VIS 1 ll U . I'fY SUN INTENSITY INTERPOLATION CROSS SECTIONS o ..... _......,.--: ,c;--\.IATERSIIEil SLOPE/ASPECT AUTOHATED CONTOURS 3-U VIEIJS DIGITAL TERRAIN ANALYSIS FUNCTIONS Figure 19 41

PAGE 49

(Section B in this chapter outlines some specific overlay examples). The various digital terrain analysis functions outline d in Figure 19 and describe d below should be of special interest to LAs. These functions are computed from a digital elevation mod el. Visibility --primarily use d i n conjunction with determining what can b e seen and .not seen from a give n x , y and z coordinate . Output is in the form of a grid o f value s indicating see n and non-seen values. Sun Intensity/Slope/Aspect --primarily involves the calculation of slope aspect, and in some cases, the relationship of this slope aspect t o solar radiation or sun intensity. Interpolation/Contouring --typically involve s the ability < to take random or regularly spaced data and generate a grid or other structure d mod e l framework on which automate d contouring can occur. Based on this interpolated model, a contouring program actually computes and subsequently draws out the contour lines. Watershe d Computation --involves the computation o f the ' watershed boundaries, drainages, topographic valleys and p eaks. In some cases, the s e d ata are used to compute surface flow. Visual D isplay --consists primarily o f cross sections and three-dime nsional views. Other GIS ma nipulations o f interest to LAs a r e listed briefly below. Most o f the s e r epre s ent combinations o r 4 2

PAGE 50

variations of the basic functions already described. It should be noted that there are many more GIS functions than those listed here; these are meant to be representative overall, with an emphasis on those which could be most useful to landscape architects. Graphic query -define an area of a map on the terminal and ask to see everything about it. Conversion Scaling/Pro portional Buffering -Windowing -Corridor Delineation -Boolean Attribute Retrieval -Statistical Summary -convert data from small units such as Douglas Fir to larger units such as ever green vegetation. size of output from 3 ' by 3' to 8-l/2" by ll" or scale of data from l" 20' to metric Delineate a 20' zone around a facility or road and analyze and display all the cost alternatives in the zone. Zoom in to a small area or step back for the big picture. Based on a number of criteria, find a corridor for a road, powerline, or linear feature. Determine all of the polygons of a certain type of soil which have an area of greater than 50 acres. For the above (boolean) example, generate a statistical summary of the polygon acreages, perimeters and their totals. 7. Output Formats and Output Devices Output products vary in format and are dependent upon software, hardware, and user preference. They include 2-D, 3-D, and 4-D maps, graphs, charts, and reports. Various GIS output are reproduced in Figure 20. Options such as color and shading also vary from system to system and may be dependent upon accompanying hardware. 43

PAGE 51

Visual Sensitivity GIS OUTPUT F igure 20 4 4

PAGE 52

The attention given to the map in a GIS should not over-shadow the importance of report generation. Beyond simple tabular listings, a report generating program can provide: a) summaries of system status {tables of attribute and locational data, size of data base, use statistics, update information, etc.), b) statistical tables, c) user guides, d) reports for presentation. Output devices are many: line printers, fixed carriage type-writers, drum and flatbed plotters, electrostatic printers, Cathode Ray Tube {CRT) terminals, computer output on microfilm {COM), color xerography, and color slides. Of these, the most commonly used are the CRT and the plotter. The CRT {Figure 3) is the TV-like screen which data as it is being manipulated. "Interactive" or graphics software allows the user to specify changes a screen image and will then make the changes instantaneously; this feature is considerably more convenient than "passive" graphics software which would have to take the time and effort to redraw the entire image.1 7 ' A CRT can be interfaced with printers or plotters for a permanent image. The plotter {see Figure 3) produces a permanent image on paper, acetate or film. A number of drafting/drawing pens can be used to produce the image: felt tip pens, technical ink pens, ball point pens, scribing tools for directly producing photographic negatives, or a point"light source for exposing 45

PAGE 53

photographic negatives. In general, line printers and fixed carriage typewriters are less expensive than plotters, but produce an inferior output, generally unsuitable for presentation maps.l8 Computer output on microfilm (COM) produces a fine resolution output which is convenient for viewing and storing maps which can be enlarged. Color slides are relatively new options for output which also provide fine resolution output. The two main issues in considering map output format are: l) for what will the output be used -for working level or for presentation, and 2) what degree of resolution or level of accuracy is needed. If presentation quality maps and a high degree of accuracy is required, the plotter is the obvious choice. If these are not needed, the line printer or fixed carriage typewriter can be relatively inexpensive alte:natives. D. Classification of Systems As mentioned earlier, GIS uses-cross over many professions and applications (see Figure l). Though many of the functions overlap, GISs in general cover a broad spectrum of focuses. GISs ' vary in design, orientation or focus, and thus in their overall capabilities. Consequently, the degree to which individual systems are appropriate for the LA profession or for specific applications varies. This section classifies GISs according to l) orientation or focus, and 2) size of the accompanying hardware. 46

PAGE 54

l. Orientation of the System Dangermond classifies GISs according to the following categories based on the purpose of their initial design.l9 The first three are of most interest to LAs. Generalized Thematic and Statistical Mapping Systems -Typically on mini or mainframe computers, these are the systems most commonly utilized by LAs. Used for natural resource management, forest inventories, vegetation, geology, soils, census mapping, environmental planning and assessment, etc. CAD/CAM Adaptations -Typically a mini-computer CAD/CAM system for engineering applications such as photogrammetry, topographic basemaps, road engineering, and utilities. These systems vary in the amount and complexity of GIS analysis which is possible. Many of them are primarily graphic systems, and consequently lack some of the analytic and location functions expected of a full-fledged geographic information system.2 0 Their main limitation for landscape architects involves overlay and composite work. Graphic overlay and composite information can be generated, but analytic comparisons and relationships in terms of specific coordinates are not usually possible. Also, if ' location identifiers are not used, important functions such as "visibility" cannot be performed. Their main advantage for LAs is their generally superior graphic output.21 Because of their focus on the design and visual elements, there is generally more focus placed on the quality and quantity of output options. It has been suggested that the functions and focuses of GISs and CAD 47

PAGE 55

systems are merging and that these once separate and distinct technologies are now taking advantage of research and development in each other' s field.22 It is important to note the distinction between the two at this point, however, to assure that when the full GIS capabilities are needed, they are guaranteed. Image Processing Systems -Usually associated with processing LANDSAT and relate d satellite image data, these are also used by LAs in l?rge scale regional projects. Property or Parcel Information Systems -Typically a mainframe based data base management system to handle attributes associated with the land parcel. Geographic Base File Systems-Most often associated with street networks and the aerial units which they define (e.g., the DIME system used by the U.S. Census Bureau). Bibliographic Systems Catalogues a variety of bibliographic data sets about geographic documents. The important thing to note here is that different functions support different levels of accuracy, different types of data ' management, different spatial identification techniques, and different approaches to information display. There is some overlap in system t y p es, but knowledge o f the basic uses for which a GIS was initially designe d can provide a goo d clue to its emphasis and can offer h e l p i n judging i t s appropriate n ess for a particular project. For e xample, i f the main purpose for 48

PAGE 56

employing a GIS is to do a statistical analysis of areas and relationships (e.g., for preparing environmental impact statements) a GIS which has been adapted from a CAD/CAM system may not have all the capabilities needed. If, however, graphical overlays will be a major emphasis with only minimal specific analysis needed, a CAD/CAM adaptation may be suitable. 2. Size of the System GISs may also be classified according to the size of the computer used with the GIS. This determines speed, the size of the data base which can be effectively handled, and resolution. Early GISs could only be run on the largest class of computers, the mainframes. As the mid-sized computers, the minis, grew in power, and as the GIS software became more efficient, GISs were developed which could handle large data bases on the minicomputer. Today, GISs are used even on micro-computers, although with some degree of trade-off in accuracy and resolution. IMGRID, a GIS developed at Harvard University in the mid 1970s is one such system.21 IMGRID has strong analytic and modeling capabilities and has been used extensively for site ' selection. Researchers found that the mainframe system could be adapted for use with the mini and ultimately with the micro-computer by reducing the data from polygon (curvilenear) to raster (grid) form. There was a significant but acceptable degree of loss in processing speed and resolution, yet there was also a radical reduction in system cost. 49

PAGE 57

The significant reduction in the cost of microcomputer geographic information system capabilities as compared with large computer systems provides an opportunity for an expanded and diverse user community including such small businesses as landscape architecture, planning, engineering firms, and universities.22 For LAs, the main considerations in this selection process are site size, scale of analysis, and degree of resolution necessary on any given project. MacDougall suggests that "most regional landscape studies will probably continue to be done on mainframes and mini-computers until there are very significant increases in the speed of micro-computers. " 23 Nevertheless, it is clear that certain kinds of analyses canbe handled quite effectively by the smaller computers accessible to small LA firms. E. Considerations for Landscape Architects Because of the unique orientation of LA work, this section summarizes the basic considerations about GISs for the landscape architecture profession by examining issues of particular concern to the profession. These issues include: advantages and disad-vantages of GISs for landscape architects; data scale and accu-racy; GIS costs, and finally, our current methods of work in relation to emerging computer technologies. ' l. Advantages The main advantages of geographic information systems for the profession include: l. Th e designer or planner can consider many more factors because the GIS can help put the information (or combinations of information) in a more comprehensible 50

PAGE 58

form. Easily accessible information includes not only environmental data, but also socio-economic and administrative information. 2. Rapid and repeated conceptual models or development scenarios can be investigated and tested against criteria established by the LA. This increases the likelihood that the optimal alternative will be identified and also allows for a more clearly defined process with quantitative results. a. It is also possible to test models with regard to changes over time. b. Client participation can be encouraged 1n the decision making process through the testing of various models. 3. It is possible to do analyses that would be or inefficient to do manually, (e.g., in-depth runoff calculations for very large areas). Use of a GIS can thus increase the accuracy and credibility of overall results. 4. Analysis data, design and management plans can be maintained in a physically compact format (i.e., the magnetic file). This reduces "paper shuffling" of maps at different scales, reports, tables, and other text. Essentially, an efficient work space/filing system is built into the system. 5. It is possible to quickly and efficiently change scales and to work on portions of a project in more or in less 51

PAGE 59

detail. 6. Editing and updating of the data base is fast and efficient. The system has built in checks for common errors. 7. Among the most useful analytic manipulations which can be done quickly and efficiently are: terrain analysis (slope, aspect, cut-fill calculations), sun/shade analysis, overlays, 3-D modeling, view shed delineations, area and frequency analysis, drainage maps indicating direction of flow, sink holes and ridge tops. 8. More effective presentation can be made through the use of sophisticated graphics output, e.g., three dimensional perspectives.showing what the site will look like after development. 9. The productivity of individuals or groups using GIS ' maps is increased because of the greater consistency, versatility, and accuracy of a GIS. 2. Disadvantages There are also a few disadvantages to using a GIS which need to be considered by landscape architects. Som e of these are: l. High initial hardware and software cost. The prices 52

PAGE 60

vary depending on the size of the system, but even a small, basic system will cost around $15,000. (See "cost" later in this section for further discussion.) 2. Time and money to convert data not already in digital form. For this reason, a GIS is most advantageous if the data are to be used more than once. Most landscape architectural offices, however, work on sites on a one time only basis. Until scanning and automatic entry technology advance to the point of substantially reducing costs, then, cost effective only smaller one-time projects may be if the data are available in digital form, or if limited data are used. Certain kinds of data require much more time to input and ma y not be needed for every project. For example, soil data are some of the most time consuming Fortunately, in many areas, Soil Conservation Survey (SCS) information is already available in digital form. For those areas where it is not available, however, all the soil data may not be required for every project. The analysis may only require generalized categories, wnich would allow ditigizing of only larger categories of data; for example, clay soils, sandy soils, etc., instead o f the more specific categories mapped by SCS. This could save a great deal of time and expense, and still be of sufficient accuracy and detail for many types of analyses. 3. Landscape architectural orientation is not built into 53

PAGE 61

3. Scale many GISs. Many GIS companies are just beginning to realize the market for the design professions. In the past GISs have tended to be built and designed for whomever could afford to fund the necessary research and development. Consequently, most initial systems were developed for utilities, engineering and geophysical exploration applications. Although these can, and haye been, adapted for LA applications,it is important to understand the possibile uses of GISs for LAs so that we are able to communicate our needs to GIS companies as they develop and modify their systems to better suit our purposes. Most landscape architects who are familiar
PAGE 62

One project was a nationwide study for the Nile River Val-ley. The scale and resolution of the project were on a gross scale typical of projects this size. All analyses were done on the microcomputer GIS, though most manipulations and mapping were performed overnight. In contrast to this project, Camarata dis-cussed smaller site specific applications using the same micro-computer GIS. For these projects, data bases were l imited t o information available_digitally or which could b e digitized fair ly quickly.24 Primary manipulations included topographic and elevation data for modelling runoff and slope analysis. 4. Accuracy One of the advantages discussed earlier in this section wa s that GISs can provide quantitative results, ensuring greater accuracy in the overall analysis. While this is a , valuable feature, it is important to understand (and to b e constantly reminded) that quantitative results can only be a s accurate as the accuracy of the original data entered. For this reason, determining the appropriate scale and detail at the outset can ensure l) that the data are appropriate for the level of analysis needed, and 2) that time is not wasted digitizing unne c essary data. This proble m was exemplified 1n the NWCCOG project which wa s discussed in S ection B of this chapter. On the site-specific analysis for studyin g dev elopmen t suitability , the set of soil data was one o f the main factors in the site s e l ection equati o n . Unfortunately, instead of collecting all the p ertinent data, a 55

PAGE 63

decision was made to use readily available data from the Soil Conservation Service (SCS). Not only were the SCS maps insufficient for this analysis because of their l :20,000 scale, they also failed to provide the necessary soil depth profile. Because of this lack of accuracy, the results were worse than useless -they were misleading. The consequences for this section of the NWCCOG project were disastrous. 5. Cost GIS software is available from private companies, some public entities, and some academic institutions. Most software from public entities (e. g., BLM, U.S. Forest, U.S. Fish and Wildlife Service) is available either free of charge or at minimal cost, while academic institutions provide software on a proprietary arrangement basis. Although the initial cost is lower, usually support services are not provided, and the software is generally less versatile, having been developed for specific "in-house" type projects.25 Frequently commercial GIS software is provided in one or more modules, so that a client need not acquire all the software that an organization offers. Such modules typically start at around $1 0, 0 0 0. An average GIS costs approximately $100,000 for all needed software modules. The hardware will typically cost the same or more.26 GIS micro-computer turnkey systems (i.e., systems with all the necessary software and hardware included) range from approximately $15,000 to $75,000 in price. At present, there are GIS systems available at the "low end" 56

PAGE 64

($15,000 to $75,000) and at the "high end" ($150,000 to $1,000,000) but not many in between.27 6. Rethinking Our Old Methods It has been suggested in this paper that computers in general and geographic information systems in particular, have the potential to restructure the form and mode of LA practice. Discussions so far have emphasized improvements ln terms of accuracy and efficiency. While this is an important change, it is only a part of the potential for change and growth. Now , more than ever, it is essential to step back and look at our old ways of doing things. Before we automatically transfer those old methods onto computer technology, we must ask if those methods are still valid. The technology offers us tremendous opportunities to view and to question • , -consequences of our decisions before taking decisive action. This has enormous implications for the LA profession, and the potential to radically alter the way we go about our business. In examining trends in the evolution o f GIS technology, a recent article in the Harvard Library of Computer Graphics asserts tha "the rapid evolution of GIS technology has not been matched by a corresponding increase in our understanding of how natural systems work. Thus, as in other areas, technology is probably outstripping our ability to apply it intelligently; our reach exceeds our grasp considerably."28 It is up to us, then, to apply the technology intelligently, to remain cognizant o f the fact that the computer does not perform feats of magic. The 57

PAGE 65

qua 1 i ty of data entered and the methods of eva 1 ua tion and analysis used will always be reflected in the quality of results and output. ' 58

PAGE 66

CHAPTER III GIS EVALUATION CRITERIA FOR LAs This section outlines a basic set o f criteria for evaluating GISs in relation to the unique set of requirements of LAs. Since software is the primary consideration, in selecting a GIS, the general thrust of this section 1s on the selection of GIS software, but a few comments are offered about basic hardware considerations as well. There are a host of GISs and related mapping software from which to choose. In order to select the system most appropriate for your purposes as a landscape architect, a four step process is suggested here to help narrow the range of GISs to ._Fonsider. The process as outlined in Figure 21 is to l) assess your specific needs and define the tasks for which the GIS will be used, 2) determine the appropriate size and orientation of the system, 3) evaluate the reliability o f the GIS vendor, and 4) appraise the GIS for ease of use and flexibility. Some considerations regarding system costs are also discussed briefly in the final portion of this chapter. A. Establish Your Needs and Determine Your Requirements Evaluating any computer system for a specific situation requires a thorough understanding of how things are currently done. The organizational structure, the physical work settings, 59

PAGE 67

1. DEFINE YOUR NEEDS 51!f t. TYf'f.5 Of MAN rMJ T5 kUNct"ION'.) useo •MtioN.l . f.QUIP\fD f'lro.IIU\W f--• t-LSO,ooo ,__ -n..v ree'-41 . • pol)',.,., \'M-he-1 • t.ite • l : w • .oy'id DETERMINE 2. SIZE AND ORIENTATION OF THE SYSTEM 3. Of'alf mltfiOJ.l Of Tlle o.M OJ'dl. • CJ.I>{ CHECK THE RELIABILITY OF THE VENDOR • Silt OF 1Jif.. • H.i.U 5 • • EVALUATE THE GIS FOR EASE OF USE AND FLEXIBILITY ; TI'T'< !Hfi. thiiJCc Yov,., OI'JU pp.a>c:;q' I • • ? • !" CRITERIA FOR EVALUATING A GIS Figure 2 1 60

PAGE 68

and the type and amount of work done a 11 need to b e examine d in detail. How will the computer change this structure? Will it expand or eliminate the work that is handled in the setting? A close examination of these issues and an establishment of goals and designs for the future will make the transition to computer use more effective and more friendly. This should be the first step in the GIS selection process. Examine how a GIS will fit into your particular situation. How will the syste m emulate or change the way things are currently done? How many people will use the system at one time? Having gained a general understanding o f geographic information systems, define the purposes for which the GIS will be employed. List the basic manipulative functions which will be used and the order of their relative importance. Determine the size and types of projects which will be studied, range o f scales which will be used, the type and amount of data to b e analyzed, and the degree of resolution and accuracy required. How quickly do you need results? What is the quality and range of graphic output that you require? Answering these questions and establishing priorities will serve to focus your search. ' B. Determine Appropriate Size and Orientation of the System The next step_ in narrowing the search for an appropriate GIS lS to determine the appropriate size and orientation of the system. This is based on the discussion of classifications of systems in Chapter II. 61

PAGE 69

l. Size of the System In general, the more demanding the requirements, of the GIS the larger and more powerful it will need to be. This is particularly true if speed of data access and manipulation is important, if large-scale projects will be undertaken, or if a large number of p eople will use the system at one time. As was discussed in the introduction to Chapter II, many o f the functions of a CAD system and those of a GIS overlap. For this reason the considerations for purchasing CAD systems and GISs are in many respects quite similar. In his article "Wary Buyer's Guide to CAD System Demonstrations," Tim Varner discusses the issue of selecting the appropriate sized system for your needs. In deciding between a mainframe based system and a less expensive mini based system, he suggests that are gray areas where the capabilities of the differently priced systems will overlap. Therefore, what you may be paying for in a more expensive system is the speed of operation. Keep this in mind when defining your needs and priori ties, to make sure the extra speed justifies the extra expense." Varner also maintains that along with increased speed of the mainframe-based systems come increased price and ofte n increased problems. A system crash on a mainframe computer, he notes, could result in several days o f down time for all of its users.29 A mainframe-based system may be justifiable only if you require high speed transactions, have a large number of users (e.g., over 25 at once)30, or if you have a mainframe computer 62

PAGE 70

already. In most cases, a mini-computer GIS will have comparable capabilities to those of a similar mainframe based GIS. The main advantage o f the micro-computer GIS is price. If your budget is tight and your needs limited and specific, a micro-computer system might be the answer. number of micro-computer GISs designed There are currently a for LA work. IRIS International, Inc., for example, has geared their software to the needs of the design and planning profession.31 It has been suggested that at least 2-4 megabytes of memory are generally needed for micro-computer GIS analysis.32 For more discussion on micro-computer GISs, see Chapter II, Section D. 2. Orientation of the System Having selected the size of the system, it is helpful at this point to narrow the selection further by deterll).).ning the orientation of the system which is appropriate for your needs. Using the classification system presented, it was suggested that the three orientations most relevant to LA work are 1) generalized thematic and statistical mapping systems, 2) CAD/CAM adaptations, and 3) image processing systems. To determine which of these systems would be best suited to your purposes, review your needs as established in the first step of this process. If, for example, it was determined that satellite data will be the primary data source, reviewing primarily image processing systems will serve to focus your search. 63

PAGE 71

C. Evaluate Reliability of the V endor A criterion of major importance for evaluating GISs is reliability of the vendor. this regard are: Some basic questions to answer 1n l) Is the software guaranteed? 2) A r e software updates provided? Free or at a p rice ? 3) Is the training of personnel included in the purchase price ? Varner asserts that "an intensive training class can reduce-by as much as three weeks-the time required for ... operators to be productive on the system." He ad vises wariness of vendors who claim such user-friendly systems that the company provides only two or three days of class room instruction and a toll free number to call if problems arise.3 3 4) How accessible are the v endors to give advice and instructions and to answer questions? 5) Who are the current users and what has been their experience with the system and with the vendors? 6) Does the vendor write the software or is it purchased from a second party?34 7) Can you receive hardware and software s ervice local l y and at what cost? Robert Puterski, senior cartographer for Colorado Department o f Local Affairs, clai ms that l ocal s e rvicin g can m ean the difference between l-3 days for local s ervicing versus weeks or e v e n months if equipment must b e sent back to the manufacturer. Th e 64

PAGE 72

industry standard for such s ervices ranges from .75 to 1.5 p ercent of the purchase price of the equipment per month. 35 8) How financially stable is the v endor? Doe s the company have the p ersonnel and finance s necessary to continue the d e v elopment and enhancement of its product? How large is its r esearch and d evelopment staff? How large is the and support organization?36 D. Appraise the GIS for Ease of Use and Flexibility This s ection provides a list o f GIS feature s which are desirable because they l) mak e the .syste m easier to use and/or 2) allow for flexibility so that the system can be update d to suit your needs and adapte d for new developments in GIS technology . These factors probably most easily discerned at a GIS demonstration. When arranging a demonstration, it is important to communicate your needs to the vendor so that s/he can tail or the demonstration to meet your specific objectives. If possible , arrange to bring in some of your own work with which to try the ' system. This is a good way to observe the syste m b eing use d "fresh." Canne d pres entations can b e misleading. 3 7 Som e features to evaluate during the d emonstration a r e : l) How user-friendly i s the syst e m ? Look for t echnique s or ope rations which seem unusually complicate d o r cumber s o me . How ofte n doe s the operator use the 6 5

PAGE 73

vendors?39 How flexible or expandable is the system? E. Cost Considerations One way to ease the high up-front cost is to purchase the software in modules, purchasing only a core system at first and adding on when additional funds become available. For LAs who will use the GIS for projects where the data will be used only once, data base updating and editing features may not be important.40 It may be desirable in this case to purchase a system without this expensive feature of the Data Base Management System (DBMS); this can sufficiently reduce required memory, cost, and overhead. The four step process outlined in this chapter is meant to clarify the most important issues concerning the evaluation and < selection of a geographic information system for a given purpose or setting. Following each step in the process and answering the listed, it is hoped, will narrow this selection to a few of the most appropriate GISs in each case. These systems can then be evaluated against one another based on the specific needs and priorities established by the potential users. ' The final evaluation should examine each system based on how well it can be tailored to the needs of the users, howclose it comes to the designated budget, and to what extent it will grow and adapt to the changing needs of the users. 67

PAGE 74

CHAPTER IV RECOMMENDATIONS TO UCD LA PROGRAM The initial recommendations made here are modeled after the previous Chapter, "Criteria for Evaluating GISs". First, the needs and requirements of the LA program at UCD are discussed. Appropriate GIS classifications for UCD are then determined, and And, suggestions are offered on how to incorporate GIS technology into the LA program curriculum. Finally, some logistics concerning staff and training are discussed and options for cooperating with public and private agencies are explored. A. Needs/Requirements of the UCD LA Program< While the primary purpose of using GIS technology is to aid 1n the design and analysis process, a second purpose exists in the unique environment of an educational institution. This second purpose is to expose students and professionals to the of GIS technology and to provide training so that it becomes a useful and comfortable tool. Since the LA curriculum at UCD encompasses the wide range of aspects of the LA profession in general, the needs of the program in the area of GIS technology are quite diverse. All the GIS functions suggested in the previous chapter as most advantageous 68

PAGE 75

to LAs in general, then, would be useful throughout the curricu-l urn. Basic modeling functions, for example, would be useful early in the design sequence, while certain functions such as overlay would be most useful later in LA 601 (Design IV) during regional design. (More suggestions regarding incorporation into the curriculum are discussed later in this section.) The size of projects and scales vary from smaller site specific to large regional projects. For smaller and site specific analysis, the higher resolution polygon data structure would be required, while for regional projects the more efficient grid structure would often be necessary to reduce memory requirements and to increase speed. Output sizes are a function of the use to which it will be put. Since most projects are used for presentations and some are used for publication or photographing, the larger input/ output devices are most desirable (e.g., a large digitizing table and a drum plotter). The number of users who would be on the system at one time is the most difficult factor to determine. The system could ultimately be used by LA students and professionals, as well as ' by other students in other divisions of the college. As suggested in the "Classifications of Systems" chapter, this is a crucial factor ln determining the size of the system. This decision should be based on l) amount of space available, 2) the commitment (money and staff) on the part of the university, and 3) interest in GIS technology from other divisions within the 69

PAGE 76

College of Design and Planning. As shown in Figure 21, the last consideration for assessing the needs of the LA Program is the speed required of the system. This is another area which should be defined by the program. Th e advantages of a higher speed system would be a faster learning curve, a lower frustration level, and increased likelihood that students would consider GIS technology as a viable tool for use on projects in a professional setting. But if the university's commitment (i.e., funding and staff) are minimal, and a micro based GIS is thus acquired, this handicap could be minimized through the careful structuring of projects and assignments. B. Appropriate System Size and Orientation for UCD The next step suggested 1n evaluating a GIS for UCD's purposes is to classify the system 1n terms of overall orientation, and in terms of its size. l. Orientation of the System The systems which would be best suited to the LA Program would probably be those most commonly used by LAs in general, ' i.e., generalized thematic and statistical mapping systems. These are the systems most often used for regional design and analysis, but they can be used for a number of other purposes as well. While not as versatile overall for the needs of the program, GIS CAD/CAM adaptations could nonetheless be integrated into 70

PAGE 77

parts of the curriculum if it is determined that purchase of a GIS per se is not feasible. This would be desirable if the College of Design and Planning purchases only a CAD system with mapping software as an adjunct component. This type of system would be most beneficial for use in first year design classes for modeling and site analysis and for some kinds of engineering work, yet would be of limited use for regional analysis and design. 2. Size of the System The process of establishing needs and determining GIS requirements will generally serve to define the size of the GIS which will be most appropriate. The particular requirements which most affect what size GIS is needed are l) number of users on the system at one time, 2) speed of transactions, and 3) budget. Since these three factors have yet to be defined by UCD, a single appropriate GIS size category cannot be established here. This section will therefore discuss advantages and disadvantages of the three GIS size options for UCD. Chapter III suggests that mainframe GISs may provide some increase in speed of transactions, but that their additional ' expense and increased problems often. do not warrant their selection over minicomputer GISs with the same fundamental capabilities. A mainframe GIS would be impractical for use at UCD for these reasons also because its larger size could not likely b e integrated into the limite d computer space currently available in the College of Design and Planning. Several space 71

PAGE 78

alternatives are currently under revie w, but it is probable that space will be at a premium until the new UCD facility is built o n the Auraria campus. Minicomputer GISs with the same basic capabilities a s mainframe GISs, but with higher r esolution, greater speed, and more available memory than micro-base d systems would likely b e the most practical and v ersati l e f o r ourpurpose s at U CD. They could be incorporated into man y aspects of the LA curriculum and could be used by othe r divisions in the Colle g e as well. Because of this diversity, they would be b ette r able to grow with the changing needs of the College of Design and Planning, would allow for more experimentation, and would be more feasible as a resource for use by outside professionals. The biggest disadvantage is' their high cost in relation to microcomputer GISs. This would require a larger university commitment in terms of both money and staff. While various funding options were previously suggested, these would be worth pursuing only if the system is used at something approaching its full potential. This would r equire staff to teach introductory computer courses, to train students on GIS use, and to organize ' and supervise a variety of GIS projects. Though not as versatile as a minicompute r GIS overall, a micro-computer GIS might b e a good introductory s ystem for UCD. While a microcomputer s ystem wou l d be conside rably l ess versatile and unable to handle as m a n y user s a t o n e time , i t has the advantage of being less expensive and more compact . The s e 7 2

PAGE 79

factors increase the likelihood that a GIS could be implemented at UCD more quickly and used sooner. A micro-based system cound be used for smaller projects on a limited basis as a way for students and professors to become conversant with the basics of GIS capabilities and the structure of GIS projects. A phasing plan might be adopted whereby a smaller, less expensive micro GIS would be purchased initially to familiarize students with some basic modeling and analysis manipulations. Then, as staff, available space, funds, and needs grow, a larger minicomputer GIS could be purchased. The micro GIS would be maintained for introductory courses and used for first year projects (see "Incorporation Into the Curricul urn"), the minicomputer system would then be used only by those students and professionals with GIS experience and for larger projects requiring more users at once, or for thesis projects where higher resolution would be necessary for in-depth analysis. UCD students would then b e trained to use both systems and would understand their potential for the variety of LA positions available to graduates of UCD (i.e., small LA offices, planning offices, public agencies, etc. ) . This option of purchasing a microcomputer GIS now and expanding to a larger minicomputer GIS later appears to b e the best option for UCD. First, it ensures earlier implementation and GIS use and second, it offers the broadest and most well rounded program overall, including as it does both microcomputer and minicomputer GIS applications. 73

PAGE 80

3. Cooperation with Public and Private Agencies To increase the of students' GIS skills and to enhance ucos image, it would be advantageous for the LA program to encourage cooperation with area professionals. Two ways o f structuring this cooperation could also help raise the necessary funds to purchase a GIS. The first strategy has been suggested by uco's architecture and interior design divisions. This was a plan devised for soliciting donations for the purchase of a CAD system for the College. In this strategy LA firms could be solicited to donate money to the program's GIS fund. Since the university has tax free status, the donations would be a tax advantage for these firms. In return for their donation, the firms wou 1 d be offered time on the compu t er, perhaps pro rated for the amoung of the donation. This time might be limited to hours or times of year when UCD student demand for GIS use is low. The second approach to encourage professional support is by way of offering short seminars geared to the needs of these public and private firms. These could be held during summer months or between semesters when GIS use by UCD students would be 1 ow. seminars which might be considered are 1) an introduction to GIS technology; what it is, how it is used, some special features of importance to landscape architects, 2) selecting a GIS for a specific application or office; what to consider, what to look for in purchasing a GIS, and 3) advanced GIS training. For agencies ln the area that already have a GIS (i.e., 74

PAGE 81

several county planning and federal governmental agencies), cooperation could be encouraged in another way --sharing data bases and providing projects for students to work on. The agencies would benefit from having input on their projects while the university would benefit from the agency support and student exposure. If extensive exchange is desired or anticipated, it would be beneficial for UCD to have a GIS which is being used by a number of agencies. UCD students would be trained on this system, have an understanding of its specific features. Interchange would be smoother and hiring UCD graduates would become particularly attractive to these agencies. 4. Incorporation Into the Curriculum Many of the GIS applications listed in Chapter I, it has been noted, are related to the work of landscape architects. These aspects of the profession are explored in the LA program at UCD-and could be integrated into the existing curriculum as GIS projects. A GIS could be used in the first year design classes for modeling and digital terrain analyses, in the second year design and the ecological analysis class for in-depth analysis and suitability studies, and in the engineering courses for predicting storm run off, calculating cut and fill, etc. Various statistical and predictive analyses could be explored in Research Methods, in theses, or in students' electives. In order to be integrate d into the curriculum efficiently, some basic GIS training would first be necessary. It is 75

PAGE 82

suggested that the college of Design and Planning offer an introductory computer class which all students would be required to take in their first year. Such a course could 1) introduce students without computer experience to some basics of operation, and encourage good user habits, 2) in traduce students to the computer facilities at UCD, including all the software packages which could be used for other classes, and 3) improve the general computer literacy of the Design and Planning students. CAD and GIS training could be included as a portion of the course so that students would be ready to work on the system in the second semester of their first year. In that semester, one site could be digitized either individually or as a class, and could be used for several classes: graphics (3-D modeling and rotation), design (view shed, sun-shade, slope, and other digital terrain analyses, "what if" queries, modeling), and (cut-fill, storm run off calculations, contour interpolation, etc. ) . Having students share the digitizing efforts, and using the same site and same data base for several classes would eliminate much of the digitizing drudgery and thus would increase the ' learning curve . More time would be spent exploring GIS capabilities. In the second year, other data input options could be explored (i.e., photogrammetry, LANDSAT data, digital data bases from area agencies, etc.), and more sophisticated analyses could be investigated by the students. GIS projects would also be 76

PAGE 83

conducive to interdisciplinary work if all Design and Planning students were given the GIS training suggested for the introductory computer course. ' 77

PAGE 84

CHAPTER V SUMMARY AND CONCLUSIONS Early GIS technology was primarily designed for and utilized by engineering, \ltilities, and geophysical companies. The applications today have expanded so that users include nearly all professions which deal with geographically based data. The technology has advanced from primarily mapping functions with crude output to include sophisticated search and retrieval, composite and comparison analysis with a variety of readable color output. Clearly, the range of opportunities for the landscape architecture profession t o utilize this technology is Yet a statewide survey o f the profession (see Appendix A) shows that only 8% of the respondents fully understand GIS uses and capabilities for landscape architects. A full 85% of all ' ,___. respondents though, s ow inter{est in learning more about them. . ' I It appears that the continue to be under-utilized ' until LAs have at least a basic understanding of geographic information systems. This paper has attempted to define GIS t echnology and its importance to the landscape architecture profession. GIS projects, and the classifications and workings o f GISs were discussed, providing an overview of the basics of GIS technology. 78

PAGE 85

From this understanding, then, considerations specific to LAs were developed, and criteria for evaluating GISs for specific applications and needs were established. While a basic understanding of the technology relative to the profession is an important beginning, real strides will b e made only when the technology it utilized by LAs in their everyday design and analysis process. This change will happen only when someone within the profession shows the usefulness of GISs for a variety of LA applications. It has been suggested in this paper that UCD can help affect this change and lead the wa y in promoting GIS use within the profession in this state. For this reason, this project, also provides some general recommendations to UCD for implementing and utilizing the technology in the LA curriculum. Much needs to be done to fully implement a geographic information system into UCD's College o f Design and Planning, yet it is hoped that this project will provide an impetus to begin this process. As a major landscape architecture university in Colorado, UCD must assume this leadership role both for the sake of UCD students who will be competing \or jobs nationwide as well as for the sake of the LA community in Colorado who need this valuable tool in facing major decisions concerning growth and change along the Front Range. The rapid changes in the computer industry are occurring in the GIS industry equally rapidly. Many of these changes will help make GISs even more viable for landscape architects. Som e of the more notable changes are: 79

PAGE 86

1) The falling costs of hardware and the explosion of affordable microcomputers. More GIS software is being written for smaller, more powerful computers making the technology more affordable to more LAs . 2) Increases in interactive graphics capabi 1 i ties. These will serve to make GISs more accessible and easier to use. 3) Increase in available digital data bases and new options for converting data into digital form. These w i 11 decrease the heavy up-front time and money spent digitizing. 4) Real time data collection. With improvements in sophisticated satellite data gathering techniques, we can work with data which is being gathered while we work. 5) Worldwide networking of information. As data bases become more accessible and widespread, networking of information among areas and professions will occur, i.acreasing interaction among professions and furthering understanding o f our world. These changes, coupled with burgeoning interest within the profession have the potential for profound changes in landscape architecture in the next few years. Clearly, the computer ' industry is affecting the way we work, the way we communicate, even the way we look at the world. It is imperative that we as landscape architects become aware of these changes and become involved in how the technology affects our work, as it inevitably will. GISs are one computer tool which can help "increase the productivity, credibility and effectiveness of landscape 80

PAGE 87

architects"4 1 It is time for us to become informed and actively involved in affecting the direction that this technology develops. ' 81

PAGE 88

CHAPTER VI APPENDICES A. Survey of the Profession A statewide survey was conducted to determine how much experience LAs in Colorado have with geographic information systems, what their level of interest in them is, and what value they perceive them to b e to the profession as a whole. A questionnaire was sent to 400 members of the American Society of Landscape Architects (ASLA) in Colorado. A brief article was published in the March, 1983 Newsletter, and the questionnaire was inserted. Of the 400 questionnaires sent, 106 were returned. A summary of the results is shown in Figure 22. While 60% of the respondents used computers 1n their workplace, only 20 % had ever used a GIS. Since micro-computers and minicomputers were by far the predominant computer system used in workplace (77%) , g eographic information systems based on micros and minis are likely be the most feasibly purchased by LAs in the future. While the level o f knowledge about GISs was low (only 8 % understood most of their uses and capabilities for LAs), interest in further information was very high (85% ). These results underscore the importance of this project and in further GIS literature and training geared toward the profession. 82

PAGE 89

n=106 Colorado Chapter Amerloa• loelely of Landeoape Arohlleota (CCAIL.A) loot 9ct te% rot 60'Yo r\= ,4 401'o ns43 40't 3ot 2J:>•t. lo-t. 5"'1. I"UU<.•-I.U!'I aooou
PAGE 90

Access Accuracy Address Algorithm Alphanumeric Assembler Auxiliary Memory Batch Processing Bit Boundary Buffer Byte Card Image Cathode Ray Tub e B. Glossary of Terms The act of fetching an item from or storing an item in any computer memory device. The degree t o which a measured value is known to approximate a given value. An identification, represented by a group of symbols, that specifies a register or computer memory location. A finite set of instructions which, if followed, accomplish a particular task. A set of characters with letters, numbers, punctuation marks and special symbols. A computer program that translates instructions written in a source language directly into machine language. Any computer memory or memories used to supplement main memory. A method in which a number of data items or transactions are c oded and collected into groups and processed sequentially. < Acronym for Binary Digit, the smallest unit of information which can be stored in the computer. General term for the division between two mapped areas. The internal portion of a data processing system which serves as an intermediate storage between two different storage or data handling systems with different access times or formats. A group of adjunct bits that are operated on as one unit. A representation in computer storage of the hole patterns of a punched card. The holes are represente d by one binary digit and the spaces are represented by the other binary digit. An electronic tube with a screen that is used in computer terminals to display input and output data. Also referre d to as a CRT. 84

PAGE 91

Cell Central Processing Unit Centroid Chain Character Compiler Coordinate Core Cursor Data Base The smallest region in a grid. The central processing unit or CPE of the computer is that portion of the computer which is used to control the components of the hardware system. The center point of a mass or polygon. A synonym for a string, e.g., "a chain of coordinates". A letter, digit or other special symbol used for the representation of information. A computer program that converts a source language into an object language. An ordered set of values that specify a location. The most accessible information s torage of a computer. A movable part of an instrument that indicates (x, y) coordinates to the machine. A set of data files organized in such a manner that retrieval and updating can be done on a selective basis and in an efficient manner. Data Structure The arrangement and interrelationship of data. Data Tablet Digitization ' Digitizer A flat tablet which will output the digital position of a pointer placed at any position o n its surface. The process of converting analogue or graphic data into digital form. Manual digi tization involves the transformation of data by an operator with o r without mechanical computer processor, while automatic digitization requires the use of an automatic device, i.e., scanner, pattern recognition, character recognition. A device which converts maps into a digital format for computer input. Direct Access Interactive systems employ direct or random access in which the access time is not r elate d to the location of the data in the compute r m emory , i .e., data does not have to be serially or s equentially searched. 85

PAGE 92

Editing Encode Field File Fixed Length Record Format FORTRAN Geocoding Geographic Base File Geographic Coordinates Gee-referencing Grid Coordinates Hardware Hard Copy Information Retrieval The detection and correction of errors. The process of applying a set of unambiguous rules to transform data from its original form to some coded representation, usually digital. A group of characters that is treated as a unit o f data. A variable number of records grouped together and treated as a main division of data. Relates to a number in which various records must contain the same number of characters. The specific arrangement of data in a record o r file . . An acronym for FORmula TRANslation, a procedureoriented computer programming language. The geographic coding of the location of data items. A coded network. A spherical coordinate system for defining the position of points on the earth. -< Planimetric coordinate system which identifies points on the surface of the earth. Systems include latitude-longitude, universal transverse mercator, stable plane coordinate and land survey systems, etc. Euclidean coordinate system in which points are described by perpendicular distances from an arbitrary origin, usually on an (x, y) axis. The physical components of a computer and its peripheral equipment. Printed or paper copy of computer output. Commonly a paper copy of the information displayed on a computer video terminal. Methods and procedures used for storing and retrieving specific data and/or reference s based on the information content of documents. 86

PAGE 93

Interactive Mode Interface Intersection Light Pen Line Printer Machine Language Allows users to directly interact with the information system to input and/or manipulate and retrieve information in a real time framework. The junction between components of a data processing system. The region containing all of the points common to two o r more regions or polygons. See also union. A device the size of a ball-point pen which is used for pointing to a location on a CRT screen. One of several types of interactive positioning devices including a mouse, joystick and tracking ball. An output device for computers which prints one line at a time. It can be used as a high speed listing or, by spacing symbols, as a plotting device. Instructions written in a understood by the computer translation. code that can be without further Magnetic Disk A computer memory device on which data is available by random access. Magnetic Tape Memory Minicomputer Natural Language Network Node Object Language Of f-line A computer memory or storage device which will store a large amount of data, but data is only accessible in a sequential search. An organization of storage units (bits, bytes) retained primarily for information retrieval. An inexpensive CPU with limited core capacity. A user-oriented language which can be used to search the computer files by operators who have no programming experience. A connected set of segments and nodes. A point which is common to two or more segments. A machine language that is output from a compiler. Processing is not directly unde r the control o f the c entral processing unit. 87

PAGE 94

On-line Processing is directly under the control of the central processing unit. All interactive systems operate on-line. Optical Char-The process by which printed characters are read acter Recog-by light sensitive devices for computer input. nition Also referred to OCR. Overlay The superimposition of one map or digital image over another of the same area in order to determine data combinations or intersections and unions. Periphial Input and output equipment used to transmit data to and receive data from a CPU. Plotter An (xr y) mechanism controlled by a computer, generally for the recording of location or spatial information. Lines are drawn as a series of vectors. Polygon Plan figure consisting of three or more vertices (points) connected by line segments or sides. Program The implementation of a procedure by the use of a computer programming language. A program consists of a set of instructions which direct the CPU in the performance of a specific task. Random Access The process of obtaining information or
PAGE 95

Time Sharing Uniform Grid Union VariableLength The concurrent use of a computer system by more than one user or program by allocating short time intervals of processing to each active user. Th e response time is usual l y so fast, that each user is given the impression that the computer's resources are totally designated to his task. Square, rectangular, or hexagonal lattice grid coordinate system for recording geographic data. The region containing all of the points in two o r more regions o r polygons . See also intersection. Relate s to a file in which the various records ma y contain a different number of characters. 89

PAGE 96

FOOTNOTES 1Julius Gy. Fabos, "Paperless Landscape Architecture: Future Landscape Journal, 10, No. l (1983), 13-18. 2"Proceedings of the Harvard Conference on Computer Graphics .anning and Design," Compute r Graphics November 1983, Newslette r of the Nationa l Compute r Graphics Association. 3Gilbert H. Castle, "Regional and Site Specific GIS 1sis", theme d elive r e d at the National ASLA Conference, tnapolis, Indiana, 1984. (unpublished paper). 4 G e o f fry Dutton , Ed. , .ion G e at i o . u r •iews, and Criteria, Vol. 2 (Cambridge : President and Fellows trvard College, 1978), p. 4. 5comarc Design Systems, Wallace McHenry Roberts and Todd, 1west Colorado Council of Gov ernments, "Draft Ar:eawide Water ity Plan for Eagle, Grand, Jackson, Pitkin, and Summit :ies Colorado" (Frisco, Colorado, 1978). 6Ibid. 7Ibid. 8Tom Elmore, Personal interview, 26 April 1984. 9Ibid. 1 0 Bruce E. MacDougall , in Ltecture (New York: Elsevier, 1983), p. 207. llJack Dangermond, raphic ,ented at various Jblished paper). 12Ibid. ,;Software Components Commonly Used in Systems," working draft of a theme professional meetings during 1982. 13Julius Gy. Fabos, "Paperless Landscape Architecture : r e Pros p e c t s ? " , J o , l 0 , No . l ( l 9 8 3 ) , p . l 6 . 1 4Bri mer Sherman, I n t eractive Syst ems, D i r ector o f e ting, personal intervi ew, 10 O c t . 198 3 . 9 0

PAGE 97

Marble, D . F., ed. for !:!.ans!lin_g_. Volumes l, Geographical 2, and 3 . Ottawa, Ontario: International Union, 1980. Monmonier, M. S. "Cartography, Mapping, and Geographic I n f o r m a t i on " , frog!. e s s i_Q H u n Q eo 7 ( 3 ) , 4 2 0 -4 2 8 , 1983. Monmonier, M. S. ts. 1982. Computer-assisted Cartography: Principles and Englewood Cliffs, N . J.: Prentice-Hall, Inc., Moody, Howard, NWCCOG. Telephone intervie w . l May 1984. Moore, P., ed. Harvard Library o f Computer Graphics. Laboratory for Computer Graphics and Spatial cambridge, MA: 1982. Vols. l-6, Analysis. Naisbitt, John. "The New American Contest: Thinking Globally and Acting Locally" (abstract). Frontiers for the 80's: !_Q s o -cl.ety of Landscape Architects, 1980. Novak, Michael. The Experience of Nothingness. New York: Harper & Row Publishers, 1970. Phillips, Richard L. "Computer Graphics in Urban and Environmental Systems," Tutorial: Computer Graphics, Beatty, John C. and Booth, Kellogg S. IEEE Computer Society (Silver Springs, MD), 1982.. Puterski, Robert, Senior Cartographer, State o f Colorado Department o f Local Affairs. Personal interviews. 21 November, 6 December 1983, and 28 March 1984. Robinson, Arthur . "His Master's Voice: Computers are Learning to ListenandTalkBack." ScienceQ.Q_, l , No.3. (1980). Rowes, Ruch A. "Perception of Perspective Block Diagrams." The American Cartographer, No. 5 , No. l (1978). ' Sherman, Brimmer, Director of Marketing, Interactive Systems. Personal interview. 10 October 1983. Sleeper, David. Conservation 1980. "Technological Choices Can be Deceptive." Foundation Lette r . Washington, D.C.: January, Sonnon, David, Systems Analyst, Computer Data Systems, Inc. Personal intervie w . 17 November 1983 . 95

PAGE 98

Starling, Robert, Ph.D. "Future GIS Developments" Talk delivered at the Geographic Information System Workshop, Western Energy and Land Use Team, Ft. Collins, Co., 7 December 1983. Starr, L., and K. Anderson. "Some Thoughts on Cartographic and Geographic Information Systems for the 1980s", Proc. Pecora VII Symposium. Falls Church: ASP, 1982. Tomlinson, R. F., ed. (Proc. UNESCO/IGU 2nd Symp. Geographic Information Systems,.Ottawa, Ont, Canada, Aug. 1972). Tomlinson, R. F., and H. W. Calkins, and P. F. Marble. S?_f Data. Vol. 13, Natura_!_ Research. Paris: The UNESCO Press, 1976. U.S., Interior, Geological Survey, Overview and USGS Activities, by R . McEwen, H. Calkins, and B. Ramey, irlR. McEwen, R. Witmer, and B. Ramsey, eds., USGS Digital Cartographic Data Standards, Circular 895-A, Reston, VA: USGS, l983a. U.S., Interior, National Park Service, Draft Environmental Impact for Upper Delaware National Scenic and Recreational River/New York Pennsylyania. Oct. 1982. Weber, W. "Geographic Information Systems -A Review and Reflections on the Future Development", International Yearbook of Cartography, XIX , pp. 119-138, 1979. ' 96