The Basics of Geographic Information Systems

Copyright ©  2008  Regents of the University of Minnesota. All rights reserved.

This publication, which provides a basic understanding of Geographic Information Systems (GIS), is one part of a three-part series. A second publication, Geographic Information Systems: A Glossary (Blinn et al. 1993), defines many of the terms associated with GIS. The last piece in the series, Introduction to Data Analysis Using Geographic Information Systems (Falbo et al. 1991), describes analysis functions in detail.

Introduction

When introduced in the 1960s, the use of Geographic Information Systems (GIS) was limited to a small number of research and applications users. Today, GIS is one of the fastest growing technologies; it has applications in public safety, natural resource management, environmental analysis, utilities, and government, and is moving quickly into many other areas. The motivation for the tremendous growth in the use of GIS is clearly linked to both the increasing demand for information and the ever-increasing ability of computer technology to provide effective, cost-efficient data processing and management capabilities.

Conceptually, a GIS can be envisioned as a stacked set of map layers, where each layer is aligned or registered to all other layers. Typically, each layer will contain a unique geographic theme or data type. These themes might include, for example, topography, soils, land-use, cadastral (land ownership) information, or infrastructure such as roads, pipelines, power lines, or sewer networks. This image of GIS is shown in Figure 1. By sharing mutual geography, all layers in the GIS can be combined or overlaid in any user-specified combination. In some cases the GIS may be defined by the type of data that the system is designed to handle. For example, the term "Land Information System" or "LIS" is often applied to a type of GIS used by counties, cities, and municipalities to manage land parcel information.

Figure 1. A conceptual model of a GIS provides a useful way to visualize GIS as a set of map layers or themes, all registered together to a common map base or geographic area. Each layer typically contains one type of data. The GIS database stores both the spatial data (where something occurs) and the attribute data (characteristics of the data) for all of the features shown on each layer.

In some of its simpler operations, GIS provides an automated version of traditional map analysis. Map overlay is probably the most common GIS function; this function has long been performed manually or optically using maps registered over light tables or via photographic techniques. In GIS, however, the number of registered map layers that can be collected and stored is theoretically infinite. The user can quickly retrieve, overlay, manipulate, and analyze any number or combination of layers. The user can then assess the results of the analysis on a computer screen or on a hard copy paper map produced by the GIS, or can summarize the results in a tabular format. However simple or complex the user's purpose, a GIS is used to access an integrated, geographically referenced database of maps that can be overlaid, combined, and analyzed to user specifications.

Because GIS is a rapidly growing technology, many different definitions of GIS now exist. A very useful definition of GIS that we will adopt in this publication is a true systems-based definition: a GIS is a computerized, integrated system used to compile, store, manipulate, and output mapped spatial data. Later in this publication, this definition will be discussed in detail. The intended audience for this publication includes GIS novices as well as those who are interested in getting started with GIS. By exploring this systems-based definition and expanding on it, you will be given a succinct, yet comprehensive view of this technology.

Our systems-based definition may be called "functional" because it is based on the functions that a GIS performs. Since each of these functions is performed on the geographic data associated with the GIS, the definition may also be called "data-centered."

Another way to understand GIS is to examine the generic "components" that the system must contain. Thus, in addition to data, a true GIS includes computer hardware, software, and users. In order to accurately be called a GIS, a system must contain all of these elements. This is a crucial distinction. Note that the acquisition of a software package or a set of computer hardware does not mean that the user has a "GIS." Similarly, merely computerizing a set of mapped data does not result in a "GIS." An actual GIS includes all of the essential elements of data, technology, and people who use the system in support of data management and analysis. You should be aware that there are several technologies that are related to GIS but are not the same as GIS. Table 1 names and defines several of these technologies.

Table 1. Several technologies are related to GIS but are not the same as GIS. Advances in computer technology to support data management and analysis have simultaneously occurred in areas other than mapped or spatial data processing and so a number of technologies that are somewhat similar to GIS have become established. These technologies include Computer Assisted Drawing and Drafting (CADD) systems, Database Management Systems (DBMS), and Automated Mapping/Facilities Management (AM/FM) systems. Each of these technologies has been developed for specific purposes, such as the use of CADD systems for the generation of architectural drawings for a new building. Yet a clear distinction can be drawn between GIS and these other technologies because GIS has evolved exclusively to manage and analyze mapped data. The ability of GIS to predicate all data processing functions on geographic location makes this technology distinct. These data processing functions include not only the storage and communication of mapped data, but also the measurement and analysis of data portrayed on maps.

Both the components required for GIS and the functional definition of GIS provide a useful way to approach this technology. After systematically examining each component and function, you can then conceptually re-integrate the individual parts into a complete whole. This approach, termed decomposition, will give you a full understanding of the system.

Before examining the individual components and functions of GIS, it will be helpful to consider the nature of maps and mapped data. A GIS works with observations or measurements that can be tied to a specific geographic location on the ground. Another common term for mapped data is spatial data. Spatial data vary with location, so the nature of the data that we collect, measure, and interpret will change as we consider various locations on the earth's surface. Observations of the earth's surface are recorded on maps to portray the spatial data in a format that is easy for humans to comprehend. Maps, as devices used to communicate the nature of spatial data, are the focus of the next section.

Maps and Map Data Handling

Maps are fundamental tools used to portray spatial or geographic data. Data shown on maps vary with location. Map data are organized, classified, and depicted in a manner chosen by the map maker to optimize the map's effectiveness in communicating the nature of the data. When we read a map, we are looking for patterns, linkages, or relationships in the map's data. In some cases, we may look at several maps simultaneously to determine relationships between geographic data depicted on those separate maps.

All maps are simplifications of the real world. The true earth is infinitely complex and it is not possible to depict on a map all of the real earth features that we might be interested in. As a simplified image of the earth, maps can be called models of reality. Though they are simplifications, these models are quite sophisticated, and the science of map making, called cartography, is a formal geographic science.

Cartographers design and produce maps using very rigid rules and guidelines. A true map must accurately show not only what the nature of the mapped variable is (in correct proportion), but also must correctly place all mapped data in their true geographic locations. The geographic location of spatial data can be classified into two types: absolute and relative. Absolute location refers to a unique and standardized place or position, while relative location defines position based on the location of other variables or phenomena. Absolute geographic position is specified using a universal coordinate system such as Latitude/Longitude or Universal Transverse Mercator(UTM) coordinates. These universal coordinate systems allow both the map maker and the map user to specify a unique and definite position for every location on both the earth and maps of the earth. This unique location, which is "tagged" to all spatial data, is critical to being able to store and analyze data in a GIS. It is this geographic tag or characteristic that distinguishes GIS as a technology that focuses on mapped data.

When we read a map, we rarely depend on absolute geographic information to assess patterns or relationships. In consulting a road map, for example, to decide on a route between Los Angeles and San Francisco, the traveler would probably not need to know the coordinates (e.g., latitude/longitude) of each city to select a travel route. Rather, the traveler would visualize the optional routes and come to some decision based on a personal interpretation of the "best" route. In selecting a route by looking at the geographical network of roads connecting the two cities, a casual map user would not seek a formal interpretation of the map. However, even a casual user would assume that all roads and other data depicted on the map are shown in their correct relative positions. Also, this type of user would assume that all mileage and other ancillary data such as the locations of rest stops were accurately reported on the map legend.

For a map to be geographically correct and correctly portray relationships between, for example, roads and cities, the cartographer must follow precise guidelines when compiling the map. For our purposes, we will simplify that set of rules to three parameters: coordinate systems, projections, and scale. Only when these three parameters are specified can a correct map be generated. In turn, an accurate GIS database depends on the availability and use of correct original map manuscripts.

These mapping guidelines prove essential when we convert traditional maps into computerized, GIS-compatible data sets. By specifying locations according to coordinate systems, a robust and accurate GIS can be established. Without a common set of geographic coordinates, layers in a GIS will not register (i.e., overlay precisely), thus precluding any type of multiple map operation such as overlaying two or more maps.

In turn, the ability to generate good location data is based on the accurate projection of the original map data. Projection is the process of transforming locations on the curved earth surface to locations on a two-dimensional plane (Figure 3). This mathematical step is performed by the cartographer to ensure geographic fidelity. While most of us never consider such a seemingly irrelevant source of error as the curvature of the earth, it is a critical consideration to the cartographer, and ultimately but perhaps unknowingly, to the conscientious map user.

Figure 3. Generally there are three classes or families of map projections. Each geometric shape is used to transform the globe (a curved surface) to a plane (the map surface). These three families are called "developable" surfaces because planes, cones, and cylinders can be "flattened" without distortion. Azimuthal or planar projections use a flat two-dimensional surface to develop the map, conical projections are transformed onto a cone wrapped around the globe which is then flattened, and the cylindrical family of maps are projected onto a cylinder wrapped around the globe. Different projections are selected to minimize specific types of distortions in distance, direction, shape and area found on all flat maps.

Few GIS users will actually perform map projection or create a map from scratch. It is more likely that users will compile data from an existing map base, suchas that provided by the U.S. Geological Survey's (USGS) National Mapping Program. Fortunately, these standard map series conform to rigid accuracy standards. Furthermore, the margin of each map contains information regarding the projection and coordinate systems that were used. Thus, USGS maps often provide a basis for the generation of other GIS data layers. As GIS developers build their system, and as users access and manipulate data from multiple maps, each must be aware of the need to standardize these parameters. Common projection and coordinate systems for all spatial data must be established if the map data in any GIS are to be fully integrated and registered.

GIS developers and users must also concern themselves with map scale. Scale is the mathematical relationship of real earth distance (ground distance) to that same distance as it is shown on a map. This relationship is often stated as a ratio of the two distances. As mentioned above, all maps are models or simplifications of reality. Maps also are reductions of reality. That is, the ratio of ground to map distance is normally much less than one. Maps to be stored in a GIS must be similar in scale if they are to be manipulated together. Maps with widely varying scales cannot be accurately combined. Thus, the user of either traditional paper maps or computerized GIS maps is fundamentally restricted by the degree of scale difference between map manuscripts. Maps with large differences in scale (e.g., 1:250,000 vs. 1:9,600) cannot be registered and overlaid without serious distortion and probable error.

A GIS converts traditional (usually paper) maps into a computer-compatible form. In order that the spatial data stored in the GIS be accurate, the map maker and the GIS user must consider issues of coordinate systems, projections, and scale. Every GIS requires that data be compatible with respect to these basic issues used in compiling mapped data.

As suggested by the example of the traveler selecting a route between two cities, often our analysis of the data shown on maps is informal. In other cases map users want specific measurements of the relationships between map features. Traditional map analysis is based on manual methods such as the use of a dot grid to measure area or a planimeter to measure distance or area. One of the most significant advantages afforded by a GIS is the ability to automate map analysis functions. That is, a GIS can be used to quantitatively assess the nature of spatial data stored in a GIS map database. These types of GIS operations are discussed in a companion publication, Introduction to Data Analysis Using Geographic Information Systems (Falbo et al. 1991).

Traditional maps suffer some disadvantages over the maps that have been entered (automated) into a GIS, as outlined in Table 2. Traditional maps are static and fixed with respect not only to the data, but also with respect to projection, scale, and coordinate system. When a user needs to integrate maps of different scales, for example, it is often necessary to re-draft the maps. Map updates may similarly require a tedious manual process. Users also face practical limits on the number of maps that can be manually overlaid, using a light table for example. Similarly, it is difficult to combine multiple map sheets together into a seamless mosaic of maps which covers an area of interest that extends to more than one map sheet or manuscript. In short, a GIS is able to overcome some of the disadvantages of traditional map management and analysis.

We have established the fact that GIS is a multi-faceted, computerized set of techniques for dealing with spatial data. The conceptual model of a GIS as a set of registered map layers is a useful way to visualize GIS. Furthermore, you can now appreciate that the accuracy of the data on any one map as well as the ability to overlay or combine multiple maps is based on cartographic parameters such as scale and coordinate system. The next two sections discuss in depth the components and functions of GIS. This decomposition of GIS into its component parts and functions will give you a basic understanding of what constitutes a GIS as well as what a GIS is used for.

Components of GIS

A common misconception is that data alone constitute a GIS. Similarly, it is sometimes assumed that someone who has purchased GIS software has acquired a "GIS." A central tenet of this publication is that a GIS is a system; it consists of more than just data and software. As a system, it exists to answer geographical or spatial questions, and the answers that it provides support decision-making. Thus, it is not enough to have data, or hardware, or software, or users; all components must be present in the system to have a true, fully-functional GIS.

There are four components of GIS: (1) data, (2) hardware, (3) software, and (4) users. As shown in Figure 5, the components must be integrated; they must be linked together and work in concert to support the management and analysis of spatial or mapped data.

Figure 5. Components of a GIS include more than just computer technology. A GIS is an integrated system of users, data, hardware, and software. Data tend to be at the center of any GIS system, while the computer components of the system support data management and analysis.

Data

GIS components are dynamic; there is rapid change in the computer industry as well as turnover of personnel involved in GIS projects. For this reason, GIS developers are often encouraged to adopt a data-centered approach. Simply stated, a data-centered approach views data as the central resource in the GIS. Though data may be shared among multiple users and multiple hardware/software environments, the data are collected and compiled by a person or organization to support the goal of that user. The other components provide the support needed to process that data.

All data in a GIS are either spatial data or attribute data. As discussed above, spatial data tells us where something occurs. Attribute data tells what occurs; it tells us the nature or characteristics of the spatial data. For example, we might describe the location of a municipal water well as a point with the coordinates "45 degrees 17 minutes 20 seconds north latitude, 94 degrees 7 minutes 48 seconds west longitude." Furthermore, we can observe and report several attributes of that well, including its depth, yield, water quality, and proximity to a pumping station. Every GIS provides the ability to store and manipulate both the spatial data and the associated attribute data.

Hardware

Computer hardware used to support GIS is a highly variable part of the overall system. Users will customize their hardware environment to best meet their own individual needs. In all cases, however, a fully-functional GIS must contain hardware to support data input, output, storage, retrieval, display, and analysis. During its infancy, GIS processing was supported only on large mainframe computers. Today, a variety of platforms support GIS processing, ranging from large mainframe computers to mini-computers to scientific work stations to personal computers. In many cases, hardware used to support other applications (e.g., payroll or accounting or digital image processing) is also used for GIS. Hardware configurations for GIS span a tremendous range in terms of start-up and acquisition costs. All systems carry associated costs for maintenance and support.

Software

Software is also a highly dynamic part of the system. Dozens of GIS software packages now exist. These systems are available on many different types of hardware platforms and come with a wide variety of functional capabilities. Public domain software is also available, though on a more limited basis than commercial software. The range in software options goes from generic turn-key systems that are ready for use "right out of the box" to customized installations designed to support specific user needs.

Given the sometimes bewildering array of choices for hardware and software, selection and use of a GIS should be approached strategically. In all cases, anyone considering a GIS software package should consider needs carefully and consult various references, including other users, vendors, and technical publications.

Users

The final component required for a true GIS is users. The term "user" may refer to any individual who will use GIS to support project or program goals, or to an entire organization that will employ GIS in support of its overall mission.

GIS users are often envisioned as hands-on computer processing people. While this is in part true, we choose to define a broader spectrum of GIS users. One classification scheme (USGS, 1988) classifies users into two groups: system users and end users.

System users are those persons who have actual hands-on use of the GIS hardware and software. These persons have advanced technical skills in the application of GIS to problem solving. System users tend to be responsible not only for the day-to-day use of the system, but also for system maintenance and upkeep.

End users are those persons who do not have actual hands-on use of the system but who do make use of the information products generated via the GIS. End users do not necessarily have to possess hands-on technical skills. However, they must be able to communicate effectively and interact with system users in order to make requests for information products, and must also understand the limitations and requirements of GIS-based processing. GIS must not be a "black box" to end users. They should always be aware of the costs and benefits of a GIS approach to problem solving in order to determine if the requests they make of GIS are reasonable and to judge the quality of the output they receive.

Yet another type of user can be called data generators. These are people who collect the raw data to be entered into the GIS. Because a GIS requires such a strict structuring of the data (refer to Maps and Map Data Handling), data generators must often predicate data collection and compilation on the demands of the GIS. Thus, these persons must be made aware that the intended use of the GIS will play a major role in the way that they collect and package data.

In sum, we can see that there is a broad spectrum of GIS users, a spectrum that extends beyond the image of all users as "hands-on" GIS experts. GIS has the capacity to affect people at all stages of the data collection, management, manipulation, and interpretation process.

Functions of GIS

Another productive way to study GIS is through our original definition: a GIS is a computerized, integrated system used to compile, store, manipulate, and output mapped data. This section examines each of these functions.

Compilation

Data compilation involves assembling all of the spatial and attribute data that are to be stored in a computerized format within the GIS. Map data with common projections, scales, and coordinate systems must be pulled together in order to establish the centralized GIS database. Data must also be examined for compatibility in terms of content and time of data collection. Ultimately, the data will be stored in a GIS according to the specific format requirements set by both the user and the chosen GIS software/hardware environment.

When all of the common data requirements are set by the GIS user, a "base map" has been established. A base map is a set of standard requirements for data. It provides accurate standards for geographic control, and also defines a model or template that is used to shape all data into a compatible form. A base map is not necessarily a map per se; rather, it is a comprehensive set of standards established and enacted to ensure quality control for the spatial and attribute data contained in the GIS.

Once the data are assembled and base map parameters are set, the user must translate the map and attribute data into computer-compatible form. This conversion process, referred to as "conversion" or "digitizing," converts paper maps into numerical digits that can be stored in the computer. Digitizing can be performed using various techniques. Scanning is one technique. Another technique is line digitizing which uses a tablet and a tracing stylus (Figure 6). Digitizing simplifies map data into sets of points, lines, or cells that can be stored in the GIS computer. Each GIS software package will impose a specific form and design on the way that these sets of points, lines, and cells are stored as digital map files.

Figure 6. A digitizer can convert paper maps into numerical digits that can be stored in the computer. The digitizing simplifies map data into sets of points, lines, and cells which are stored as digital map files.

Digitization is a simplification process that converts all spatial data to either a point (e.g., a well), a line (e.g., a stream), a polygon formed by a closed, complex line (e.g., a lake), or a grid cell. Digitization reduces all spatial entities to these simple forms because they are easy to store in the computer. A GIS database cannot readily recognize features or entities as human map users do. For example, we cannot enter the entity "lake" into a GIS. Rather, we enter the spatial data coordinates for the lake's shoreline as a polygon. Later, the attributes of the lake will be entered into the GIS database and will be associated with the polygon.

Following the digitization of map features, the user completes the compilation phase by relating all spatial features to their respective attributes, and by cleaning up and correcting errors introduced as a result of the data conversion process. The end result of compilation is a set of digital files, each accurately representing all of the spatial and attribute data of interest contained on the original map manuscripts. These digital files contain geographic coordinates for spatial objects (points, lines, polygons, and cells) that represent mapped features. Although we conceptualize the GIS as a set of registered map layers, the GIS actually stores these data at a much more primitive level.

Storage

Once the data have been digitally compiled, digital map files in the GIS are stored on magnetic or other (e.g., optical) digital media. Again, different GIS software packages will employ different storage formats. In most cases, however, data storage will be based on a generic data model that is used to convert map data into a digital form. The two most common types of data models are raster and vector. Both types are used to simplify the data shown on a map into a more basic form that can be easily and efficiently stored in the computer.

Raster

Raster approaches to storing map data in a GIS are perhaps the most intuitive. Figure 7 shows the essential steps involved in converting a map to a raster format. First, a gridded matrix is registered to and overlaid on the original map manuscript. Location in the grid is defined by the row and column coordinates of each cell. To encode the map data for each cell in the raster format, three pieces of data are recorded: the row coordinate, the column coordinate, and the attribute. Thus a triplet of data is recorded for each cell in the array, which is termed a raster. After map data are stored in a raster format, each cell in the raster corresponds to a location on the map and each cell's location in the raster is identified by row and column coordinates. By assigning a value to each cell, the corresponding attribute data for that location are also stored. The end result of this conversion process is a set of cells, each with a specified location and an attribute value. These data can then be entered into a computer-compatible file and stored in the GIS database.

Figure 7. A raster model stores spatial and attribute data together for sets of grid cells registered to the original map. The location of each grid cell is defined by row and column coordinates with an associated attribute tag linked to each cell. The end result of converting a map to a raster data file is a GIS-compatible digital data file.

Perhaps the most critical issue in using a raster GIS is the selection of an appropriate grid cell size. The user is forced to examine the trade-off between data resolution (how small grid cells are in the raster array) and storage requirements (increasing the number of rows and columns causes exponential increases in storage requirements). The use of smaller cells records greater detail in the GIS, so the user would normally attempt to select the smallest practical cell size. The choice of cell size depends on many factors, including the resolution of the original map data, the degree of resolution needed in the GIS analysis, the time and money available for data compilation, available storage space on the GIS computer, and cell sizes already employed for previously existing raster data that the user may want to incorporate into the GIS database.

Cell size is critical because it is one of the base map parameters that must be standard for all of the layers in a raster GIS. While it is technically feasible to store data at different cell sizes, all analysis and multi-map manipulation operations require that cell sizes be the same (as well as projection, scale, and coordinate system). Typically, a GIS allows the user to adjust cell sizes within the database. The user must be aware of the nature of the data and not violate the limits of those data. For example, when attempting to combine two data layers that have been compiled and stored at two cell sizes, the user must increase the cell size on the layer having smaller cells to match that of the layer using larger cells. For this reason, the highest level of detail stored in such a database is only equal to the map layer having the largest cell size. This cell size becomes, in essence, the lowest common geography for the raster-based GIS.

Vector

A vector data structure is very different from a raster data model. Whereas the raster data model uses sets of grid cells to record all data, a vector model stores all spatial data as either a point, line, or polygon. These three types of spatial data are referred to as features, and a vector GIS can be termed a "feature-based technology." Figure 8 shows an example of a vector data model. When a vector model represents an entity as a point (e.g., a well), a single coordinate pair is used to specify its location. A feature represented as a line (e.g., a stream) uses a linked set of coordinates, and a feature represented as a polygon, which is an alternative form of a line (e.g., a lake), must have the same beginning and ending point coordinates. In a raster model, a point is a single cell, a line is a linked set of cells, and a polygon is a group or neighborhood of similarly encoded cells. For all three types of features stored in a vector GIS, an attribute code is entered into the GIS files to identify the object. For example, a lake would be vectorized (added to the database as a vector map) as a polygon by storing the linked set of coordinates for the shoreline and the attribute code "lake" would be entered in the GIS files to identify that group of cells.

Figure 8. A vector model stores all spatial data as primitive features or points, lines, and polygons. Points are stored as discrete coordinate pairs, lines as linked sets of coordinates, and polygons as areas with an alternative form of a line that has the same beginning and ending coordinates. Attribute tags such as "lake" are then added to the spatial data file. The end result of converting a map to a vector data file is a GIS-compatible digital data file.

Advantages of Digitized Storage

Whether a GIS is raster- or vector-based, spatial data from map manuscripts must be compiled and stored in a simplified form that the computer can recognize. Each of these models handles spatial and attribute data in such a way as to allow the digital processing capability of a computer to be applied to managing and storing data. Computers cannot store maps in their original form; they must be converted via a digitizing process. That digitizing and storage process often is a time-consuming and expensive part of building a GIS; the cost of initially digitizing all data may comprise a majority of the total cost of system implementation. Despite the cost, digital maps stored on a computer are in a much more dynamic form and can be easily and rapidly processed. Digital data can also be exchanged between GIS systems, copies can easily be made and distributed to multiple users, and data can be shared using telecommunications options available to modern computer users. Thus, the end result of converting data into GIS-compatible format is a more dynamic, flexible data environment.

Manipulation

Once data are stored in a GIS, many retrieval, analysis, and output options are available to users. These functions are often available in the form of "toolkits." A toolkit is a set of generic functions that a GIS user can employ to manipulate and analyze geographic data. Toolkits provide processing functions such as data retrieval, measuring area and perimeter, overlaying maps, performing map algebra, and reclassifying map data. A GIS usually includes a basic set of computer programs or "tools." The functions provided by the toolkit vary with the software package. Figures 9 and 10 provide an overview of various tool functions. A companion piece, Introduction to Data Analysis Using Geographic Information Systems (Falbo et al. 1991), describes analysis functions in detail.

Figure 9. Data manipulation tools include coordinate change, projection, and edge matching, which allow a GIS to reconcile irregularities between map layers or adjacent map sheets called "tiles." Query and windowing are spatial retrieval tools. Query provides a way to retrieve user-specified data from the database. Windowing allows the user to select a specified area from a map displayed on the monitor to examine it in greater detail.

Figure 10. Data analysis tools include aggregation, classification, measurement, overlay, buffering, networks, and map algebra. Aggregation helps the user in interpreting the data, classification allows the user to classify areas within a map, and measurement can be used to determine the size of any area. The overlay function allows the user to "stack" map layers on one another. Buffering examines an area that surrounds a feature of interest such as a point. Network functions examine the movement of objects along an interconnected pathway (e.g., traffic flow along a map of highway segments). Map algebra utilities allow the user to specify mathematical relationships between map layers.

Output

The final functional task of a GIS is to generate output; usually a map. GIS-generated maps are compiled from the many data sets contained in the digital GIS and match exact user specifications. Map output may employ several color and symbology schemes, and will be sized and scaled to meet user needs. These output products resemble hand-drafted maps and fulfill essentially the same purposes. However, it is incorrect to refer to GIS simply as a mapping system. Although GIS is able to generate high-quality map output, its ability to perform analysis and management sets it apart from the more limited computer-mapping packages.

Another form of output from a GIS is tabular or report information. Data summarized according to user-defined classes or within user-defined areas can readily be generated in a textual format. This output may also be routed to another computer application such as a statistical analysis package or a graphing package for subsequent analysis and display.

The digital data themselves are often overlooked as a type of GIS output. Data files can be readily shared between users or systems. Because the data are in a digital format they can easily be copied, transmitted via cable or phone line, or distributed on media such as diskettes, computer-compatible tapes, or optical media such as compact disks. This greatly facilitates data sharing and provides increased access to data and information across the entire user spectrum.

Having completed a review of the basic elements and functions of a GIS, you can understand what GIS is designed to do. A few of the relative advantages and disadvantages of a GIS approach can now be assessed from a foundation of understanding. On that basis, the next section examines a generic approach to getting started with GIS.

Planning for GIS

Because of the advantages afforded by GIS, its use is increasing rapidly. Many different types of organizations and users are interested in getting started with this new and exciting technology. Yet the developers of new GIS projects often find that selecting and implementing a GIS solution is a complex and demanding task. In all cases, new and existing GIS users are advised to engage in a strategic planning process prior to system acquisition, start-up, or the initiation of a new project on an existing system. Private consultants and vendors can assist with the planning process. It is also recommended that organizations planning to adopt GIS discuss the utility of the technology with existing GIS users who are performing similar tasks.

The concept of a strategic process is diagrammed in Figure 11. This process is compatible with the data-centered approaches outlined in Components of GIS and Functions of GIS, and follows a common-sense approach to getting started.

Figure 11. The concept of a GIS life cycle is useful for understanding a strategic approach to getting started with GIS. Users maintain a data-centered approach while designing and implementing a GIS solution that is based on a systematic evaluation of needs as surveyed during the initial planning phase. Implementation is often incremental and follows a pilot project or prototype used to examine the feasibility of a GIS for the purpose at hand. GIS systems as well as data require ongoing commitments in order to meet the GIS design goals and to maintain data that have full utility. See Somers (1990) for more detail on the life cycle approach.

Phase 1—Planning

A planning process is the first stage in the life cycle. This phase involves a systematic review of users, their data, and their information needs. This is the time to educate decision makers about the costs and benefits of GIS and to include potential users in the planning process so that they receive an overview of the technology. Once a thorough understanding of user needs is established, the second life cycle phase, system design, can begin.

Phase 2—System Design

The design phase matches user needs to GIS functionality. Design includes not only selection of hardware and software, but also the design of the GIS spatial and attribute database. Part of that database design will include base map specifications for scale, projection, and coordinate systems. Data should be tracked using a "data dictionary." A data dictionary catalogs all data entered into the integrated GIS database and maintains current records on the source, accuracy, the time of data collection, and overall nature of each and every data element stored in the GIS. Establishment and maintenance of a robust data dictionary are essential to the overall utility of any GIS.

During the design phase an incremental plan is often anticipated for implementation of the technology. Incremental implementation means that users will build a GIS piece-by-piece. Oftentimes a prototype or pilot project is implemented so that lessons learned can be used to streamline the development of a fully-implemented system.

Phase 3—Implementation

During the implementation phase, attention to all user needs must be provided through training and education. Hands-on users must be trained to utilize and maintain the system and the database. All types of users should be made cognizant of how the GIS will affect them and their data processing tasks. They must also be made aware of the changes that GIS will introduce in the area of information generation and decision making. For further information about designing education and training programs for users, refer to Conducting a learner needs assessment: A key to successful GIS development (Blinn et al. 1992).

Phase 4—Maintenance

Finally, a GIS application must be maintained and kept current in terms of data and user support. In some cases, a GIS is designed to meet the needs of a specific, finite project. In other instances, GIS is used to support an on-going mission or program. In the former case, the GIS application will terminate once the project is completed and maintenance will probably not be an issue. However, even if the initial GIS application is no longer being utilized, the data generated for the initial project may be useful to other projects or users. In those instances, a current data dictionary will be vital for determining the utility of the existing digital data for other uses.

In the case of an on-going GIS effort the system must be kept up-to-date in order to fulfill its design goals. Maintenance includes updating hardware and software, adding new data and updating existing data records, and keeping users current in terms of system functionality.

Implications of GIS

In any situation where it is applied, GIS may fundamentally change the way spatial data are managed and analyzed. When potential users are unaware of how GIS will affect them and their duties, the changes may be disturbing. GIS developers have a responsibility to initiate and maintain a focus on user issues and concerns. Ultimately, the success of a GIS venture is determined by the ability of the system to fulfill design goals. Those design goals are determined in the planning phase of the system life cycle, during which an understanding of data issues and user needs must be paramount. Potential users must be made cognizant of the role that GIS will play in their jobs, and should know that the responsibility for the success of the GIS effort ultimately resides on how well the system responds to their needs and demands.

An organization adopting GIS will need to make adjustments in user access to, and control of, information. Because of the dynamic nature of GIS, it is often difficult to track data and maintain up-to-date records on data status. Data can be copied, shared, and modified with ease in automated systems. Not all data transactions can be monitored; user access must often be restricted so that the integrity of the system is not jeopardized. Indeed, system security is a significant issue, particularly in a networked computing environment. A current data dictionary is very helpful in monitoring data integrity.

The organization must also adapt to changes in the cost of information. Because GIS requires such a strict structuring of spatial and attribute data, system developers and users must be prepared to deal with the overhead costs associated with a GIS approach. As mentioned previously, the process of converting existing mapped data into a computer-compatible format is time-consuming and costly. Conversely, a very robust data environment can be established via GIS when proper stipulations and standards are imposed on data development and use. In a GIS, data can be shared efficiently and effectively, and formerly diverse and incompatible databases can be integrated.

The GIS that results from development efforts is ultimately only as good as the data contained in the system. A GIS will not fundamentally change the nature of spatial and attribute data, nor will it correct basic inaccuracies in data. GIS provides a rigorous structure for spatial and attribute data, but bad or incorrect data of both types can be made to fit a base map standard. The bottom line is, just because data are in a GIS does not mean that they are correct.

Users should be made aware of the limitations of the GIS approach. For example, a typical GIS can technically adjust the scale of any map to user specifications. It is up to the user not to violate the precision of the data, even if the software will allow this function to be performed. For example, if data are mapped at a relatively small scale of 1:250,000, the user must decide if rescaling the map to a larger scale of 1:15,840 presumes more precision or accuracy than really exists in the data.

The powerful output capabilities afforded by GIS in producing maps or reports on paper or as displays on computer monitors provide unparalleled opportunities for visualizing spatial data. Users who generate these products must exercise care in presenting information accurately and honestly. Persons who produce information from a GIS should be aware of the intended audience and the type of information that needs to be communicated. Similarly, users of GIS output should exercise care in interpreting and drawing conclusions from those information products. In every case, knowing something about the data is critical to proper and complete use of GIS output.

Conclusion

GIS has been defined as a computerized system for compiling, storing, manipulating, and outputting spatial data. A true GIS was shown to comprise data, hardware, software, and users. A good way to analyze the GIS approach is to focus on the spatial data that is handled by the GIS.

A GIS provides a hybrid database that contains both spatial and attribute data. Spatial data describe the location of objects with respect to a known coordinate system, while attribute data describe the nature or characteristics of spatial phenomena. Data flow through a GIS via a series of system functions. Data can be manipulated in a GIS using a series of functional programs called "tools" that operate on mapped data. Users also can elect a series of output options for communicating the results of their analyses to end users.

Getting started with GIS can be a daunting task. Although the concept of a GIS as a stacked series of map data layers is quite intuitive, the actual development of a useful system demands much from system developers. As a GIS automates or computerizes traditional mapped data, the GIS user must bear the costs associated with data compilation. GIS requires that data be tied to common geographic parameters or rules called a "base map." System developers should follow a life cycle approach to establishing GIS projects or applications, which is a strategic approach that follows through on planning, designing, implementing, and maintaining these types of systems. A life cycle approach explicitly considers users' needs and also focuses on the data that must be managed and analyzed by the system.

GIS will cause profound changes in the way we collect, store, analyze, and share geographic data. Many of these changes are predictable and their impacts can be estimated before the change actually occurs. In some instances, pilot or prototype projects will be conducted to test the feasibility of a GIS solution. Users of GIS can rely on the experiences of other users and organizations as well as advice from private consultants to help them through the start-up phases of a GIS effort. Ultimately, each GIS is unique; its data, users, and applications must be tailored to fit the demands of its users.

Suggested References

Aronoff, S. 1989. Geographic information systems: A management perspective. WDL Publications, Ottawa, Canada. 294 p.

Blinn, C.R., L.P. Queen, L.R. Hegstad, and D.J. Fitzpatrick. 1992. Conducting a learner needs assessment: A key to successful GIS development. Urban and Regional Information Systems Assoc. J. 4(2):59-67.

*Blinn, C.R., L.P. Queen, and L.W. Maki. 1993. Geographic information systems: A glossary. Univ. of Minnesota, Minnesota Extension Service, NR-FO-6097. 12 p.

Congalton, R.G., and K. Green. 1992. The ABC's of GIS: An introduction to geographic information systems. J. of Forestry 90(11):13-20.

Star, J., and J. Estes. 1990. Geographic information systems: An introduction. Prentice Hall, Englewood Cliffs, NJ. 303 p.

Dangermond, J. 1986. The software toolbox approach to meeting the user's need for GIS analysis. pp. 66-75. In Geographic information systems workshop proceedings, April 1-4, 1986, Atlanta, GA. Amer. Soc. for Photogrammetry and Remote Sensing, Falls Church, VA. 426 p.

*Falbo, D.L., L.P. Queen, and C.R. Blinn. 1991. Introduction to data analysis using geographic information systems. Univ. of Minnesota, Minnesota Extension Service, FO-5740. 12 p.

Somers, R. 1990. Management topics in GIS development. URISA/GIS '90 Workshop, Vancouver, B.C., Canada. 136 p.

USGS. 1988. A process for evaluating geographic information systems. U.S. Geological Survey open file report 88-105. 22 p.

*Note: This publication (The Basics of Geographic Information Systems) and the two publications marked with an asterisk may be ordered individually or as a series. To order the series, use the title Geographic Information Systems Basics, Terminology, and Analysis Tools: A Three-Part Series (item no. PC-6136). For ordering information on these three publications, contact the University of Minnesota Extension Store (612) 625-8173.

LLoyd P. Queen is a research associate (Internet address: lqueen@mercury.forestry.umn.edu) and Charles R. Blinn is an extension specialist/associate professor (Internet address: cblinn@mercury.forestry.umn.edu). Both are in the Department of Forest Resources, University of Minnesota, St. Paul, MN 55108.

Editor/Product Manager

Richard Sherman

Graphic Design

Deb Thayer

Illustrations

Kiehne Graphics

This contribution was supported by the College of Natural Resources and the University of Minnesota Agricultural Experiment Station under Projects MN 42-40 and MN 40-16 and the Minnesota Extension Service (including funds from the Renewable Resources Extension Act).

Produced by Communication and Educational Technology Services, University of Minnesota Extension.

In accordance with the Americans with Disabilities Act, this material is available in alternative formats upon request. Please contact your University of Minnesota Extension office or the Distribution Center at (800) 876-8636.

University of Minnesota Extension is committed to the policy that all persons shall have equal access to its programs, facilities, and employment without regard to race, color, creed, religion, national origin, sex, age, marital status, disability, public assistance status, veteran status, or sexual orientation.