The Nature of Geographic Information Systems
By Charles Convis, ESRI, Sept 28, 1996
Table of Contents
What is GIS?
Top 10 Reasons Biologists Donít Use GIS
GIS and Antique Cars
Cassiniís Planisphere, a Framework for the World
What is GIS Again?
GIS and a New Framework for Conservation Biology
This is a paper on Geographic Information Systems (GIS) in the context of nature and conservation. It will describe what GIS is, how it works, and how it impacts science and society in the face of the environmental crisis.
First of all, GIS is big. GIS is not some odd little piece of software for adding maps to your clip art, although that is how many people first encounter it. GIS is instead an entire family of highly advanced computer technologies dating back almost 30 years, with scientific and theoretical orgins going back for centuries. GIS is also a multi-billion dollar industry which is appearing regularly on top ten lists of technologies for the future. GIS is a global community of users and researchers numbering in the millions, building maps and databases in the hundreds of thousands. The ESRI Conservation Program reaches more non-profit conservation and biodiversity organizations than any other technology transfer program in the world.
Next, GIS is integrative. It is a combination of organizations, theories, methods and tools uniquely powerful at integrating. GIS integrates extremely diverse data, various tools, and different ideas and people into common frameworks for analysis, cooperation and decisionmaking. Before any cooperation is possible, however, all sides in an issue must be equal. Conservation and biodiversity arguments often fail because they canít take advantage of the same technologies and data that their well-funded opponents can. Industries generally pursue the maximum profit at the minimum cost and are seldom interested in long-term anything. The only way to control reckless environmental damage is to gain enough power, money, votes, data and technology to be able to meet them on an equal footing. Industries didnít reduce pollution because of science, they reduced because a nationwide environmental movement in the 1970ís got laws into place that forced them to reduce. It wasnít always the best science, but it was all we had at the time and it made a huge difference. Biologists must continue to be engaged with conservationists, and we have to try hard to make sure that conservation advocates have the data and tools they need to be able to argue and act on behalf of nature. This also means that as voters we need to be politically active where neccessary to ensure that environmental candidates are voted into office to represent us, and we need to fight where neccessary to defeat candidates and policies that advocate the destruction of environmental protections, forests and parks.
Finally, GIS is challenging. Given the first two characteristics it would have to be. Many people and organizations have tried to establish GIS systems in the past and some have failed. Over 30 years we have learned a great deal as an industry about how to aviod failure through planning and design. Millions of people use GIS in many different guises today with complete success. The Environmental Crisis is also challenging, and we hope that as an industry we can work more closely with you in applying our knowledge and skills.
Lets take a look at how biologists have responded to this challenging new technology, or not, as the case may be.
Top 10 Reasons Biologists Donít Use GIS
1. COST: It costs too much
2. TIME: I donít have time
3. DATA: Basic data I could link do doesnít exist yet
4. CREDIT: I tried sharing data before and I got screwed
5. ONLY YOU: Iím the only one who can interpret my data
6. ITíS PERSONAL: People should just come see me
7. IíM SUPERNETMAN!: Iím putting it on the Internet, I donít need GIS
8. LETíS PRETEND: But I already use a GIS, donít I?
9. THE TOWERING DISASTER: We tried it once and it was a disaster
10. LUDDITE: GIS is just a technoweenie fad
1. COST: It costs too much This is a real concern but misses the distinction between GIS production software, which is expensive, and the more recent GIS consumer-level products like ArcView and GIS Web browsers, which take little commitment. This statement also reflects the experience of those who dive into large GIS production projects with no effort devoted to planning or design, who foolishly allow their decisions to be guided by hardware or software alone rather than by the formally-defined needs and capacities of their users. A great deal of time and money has been wasted this way in prior years. People whose impressions of GIS are based on 10 year old experiences have missed the most important developments, including the ESRI Conservation Program which focusses exclusively on making GIS technology and skills available to non-profit organizations with little or no money.
2. TIME: I donít have time While ArcView or a GIS Web site can be used at a browsing level with almost no training, it typically takes a year of dedicated effort to get a GIS database construction program underway. Another area of concern is that once existing collections and biodiveristy data begin to be mapped and related to other databases, serious quality and consistency problems are found and it takes additional substantial effort to clean them up. As professional level Conservation-GIS experts and service organizations become more numerous there will be less need for individual institutions to handcraft each program and more availablity of pre-packaged specialty products, services and data.
3. DATA: Basic data I could link do doesnít exist yet This statement reflects the generally crude state of basic, standardized global or even national biological data sets. Biologists tend to study what they can afford to get at, like big animals near roads. Microorganisms, invertebrates and remote areas are all poorly known. Moreover, even the sorts of generally needed standard databases or basemaps, such as plant communities, detailed topography, detailed protected area boundaries doesnít yet exist at the national level. Although there has been progress at the global level with products like the Digital Chart of the World and ArcAtlas, and at the national level with programs like GAP, much remains to be done and tools are needed that let us manage existing data around all those holes.
4. CREDIT: I tried sharing data before and I got screwed This reflects some earlier experiences in conservation databases where authors and original researchers were not credited or honored for the significant contributions they made, then or now, and in some cases their data was misapplied to produce garbage results. This also reflects the proprietary interest scientists often need to have in their data, which in some cases may be a lucrative source of income and so they are loath to pass it out. Conservation groups simply have to learn to get better and act more professionally in their relationships with original researchers and data providers. As a company, ESRI already understands this aspect of client relationships and we hope by exposition and example to teach this to our many conservation and research users. ESRI has numerous successful publishing partnerships for proprietary and commercial databases, but we hope that once we as a community prove that we can be more responsible in our relationships that you as data providers will find it easier to adopt the kind of open sharing critical to the success of biodiversity protection. After all, as a group, systematics organizations are sitting atop the largest, oldest and most authoritative biodiversity data on earth.
5. ONLY YOU: Iím the only one who can interpret my data Scientific databases built to support centralized research often have this problem. There were created just for internal uses and never intended to be shared or distributed outside the organization. This also reflects the often monolithic form that researchers use to keep their data in, such as large single database files, or even using a word processor as the primary software to organize their field or specimen data into files. A major part of database design is finding the relationships and structure within oneís data and setting up a data management design that reflects those relationships and manages them effectively. This canít be done with a word processor, only with a database managment system or a GIS. This is as much an institutional mission question as anything else, and weíll see later how it specifically arises from drive to solve problems for oneself rather than solving problems for others.
6. ITíS PERSONAL: People should just come see me This reflects the notion that a GIS is not neccessary since anyone who needs to use my data can just come see me in person and I can determine if I want to help them and also explain how to use my data directly. Conservation activists often share this view as well, trusting in personal contact and traditional advocacy and legal techniques to carry out their mission. What this misses is the fact that land changes are more often the cumulative result of millions of local decisions made by local planning departments, farmers and foresters. It is a fact that planning departments are more and more relying upon standardized GIS databases for the efficient conduct of their work, and using time-consuming manual or personal sources less and less. If biologists donít get into that game and learn how to get their data integrated into the forms planners use, their work will have less and less effect just when it needs to be having more and more.
7. IíM SUPERNETMAN!: Iím putting it on the Internet, I donít need GIS This often reflects the misconception that all I need to do is dump my data where everyone can see it and theyíll automatically be able to figure out all the details of how to apply it to their problem. As mentioned above, planners for one are unwilling to spend time trying to translate someone elses data. They are looking for fast answers and if data isnít already set up for the way they work they just wonít use it. As we will see, this is why the idea of a framework is so important as a means of integrating data providers and users. Without a framework of some sort, like GIS, the problems of managing distributed databases for consistency, quality and error quickly become insurmountable.
8. LETíS PRETEND: But I already use a GIS, donít I? This reflects the common misunderstanding that a GIS is just mapping graphics software the same way that a word processor lets me automate document production. It also reflects the widespread misuse of the term ĎGISí to apply to CAD-based or graphics software that really is nothing more than a drawing program with a few enhancements for graphic attributes. People may justifiably feel they donít need GIS because they donít need to produce pretty maps, but as we will see, GIS is bigger than software, and has applications in just about every area of biological sciences.
9. THE TOWERING DISASTER: We tried it once and it was a disaster A lot of organizations have tried to develop their own custom database management programs or in-house GIS software. Good custom programmers have a lot of value and strengths, but they usually lack an affinity or understanding of biological issues, producing elegantly-crafted software that rapidly and beautifully does tasks of no use whatosever to biologists. A lot of organizations find it better to let a biologist do the programming, which may be crude but will at least do the tasks biologists are interested in. Unfortunately, this all misses the point. Biologists are an extremely rare and precious breed and we need every one of them doing as much biology as possible. Let the professional software engineers create software. Commercial software is always better than a home-brew solution, which is why it costs thousands of dollars more, but since the ESRI Conservation Program donates everying ESRI makes, you can have your cake and eat it too. At ESRI, one of the fundamental jobs of a consultant, besides project design, is to mediate between clients and the applications developers to ensure that clientspeak is properly translated into geekspeak and vice-versa.
10. LUDDITE: GIS is just a technoweenie fad This reflects the view that GIS is just the latest fad among techno-weenies who wouldnít know a plant press or a study skin if it bit them in the a**. This is the justifiable view of many scientists who have a healthy skepticism about any new technology that takes too much time to manage, and diverts resources away from oneís primary mission: the direct conduct of science. As a classically-trained field naturalist, I still believe that the old, traditional, reliable, manual Humbolt & Leopold methods I learned in school are still all I need today. I still believe its a crime for a skilled biologist to turn away from important field work to doodle around with computers. I think itís the job of the industry to produce tools that field biologists can utilize immediately, adapting to their way of thinking and working rather than vice versa. Paradoxically, after years of increasing technological complexity creating decreased reliability and more complex products with impossible learning curves, the recent leaps in chip and operating system complexity have made computers a lot easier and more reliable.
When computers work well they are like driving my van, I just get in and go and I donít even have to think about how the machinery works to get me there. Some parts of the computer industry are getting close to this level of simplicity but there is still a lot of work that remains to be done.
The way computers and GIS have changed over the last 30 years is a lot like how cars have changed in the last 100. When horseless cairriages first appeared, they werenít really horseless at all since the reliability and roads were so bad a people often dragged the horse along to pull them home when they broke down! What you wanted fixed you did yourself. You didnít get an auto unless you were prepared to sink a lot of time and resources into itís upkeep.
Nowadays about all you need to buy and drive a car is a pulse, and even then I sometimes wonder. There is a national infrastructure of highways and a commercial infrastructure of motels, service stations and repair shops on nearly every block. Safety, reliability and efficiency have finally been vastly improved (after our bizarre little national machofest in the 60ís and 70ís).
Computers in the early days were also not for the faint of heart. You didnít own one unless you were prepared to spend a lot of time in binary and assembler programming. Most companies had no choice but to hire full time hardware staff and systems staff just to keep up with basic tasks. Even then, the hardware failed and disks crashed regularly. I once told a class that hardware failure was more certain than death and more frequent than taxes. If you wanted to use specialty devices like plotters, color printers or digitizers you had to write your own custom programs to run them. If you needed a basemap you had to build it from scratch, and whatever data you needed you had to go out and collect yourself.
Now, increasing numbers of standard basemaps at different scales are cheap and readily available. The web is an extremely exciting development for communication and information access, and well deserves itís place at the top of the national information infrastructure. Thousands of companies are selling specialty data and the US government is getting into the act of free standardized data. Hardware reliability has improved greatly, and with Macs and Windows 95, hardware configuration is becoming a thing of the past as "plug-and-play" becomes the standard. Finally, a huge aftermarket of custom services, bureaus, programming and hardware offers to meet any specialty computer needs for a price. We are increasingly able to think of computers as an appliance that just works and does what we want without our having to worry about how.
This is a diagram of The Environmental Systems Research Institute (ESRI). ESRI began the commercial GIS industry when it was founded in 1969. It is the largest GIS company in the world, it produces the largest market share of GIS software, and each year operates the worldís largest gathering of GIS researchers and professionals. ESRI is based in Redlands, CA, and consists of software developers and consulting staff working in a campus-like environment to produce GIS software tools and database products. About half of the institute is devoted to software development and half to applications consulting and client services. Services range from turnkey development of GIS centers at the international level to on-site consulting. Clients include national governments, states and cities, utility companies, corporations, and non-profit organizations. The data publishing program manages design and content standards for GIS databases and currently distributes several hundred titles, as well as overseeing data publication partnerships with government and non-government data providers. The software tools range from production GIS to corporate data processing to data publication and internet servers.
ESRI also operates several programs in support of non-profit, public and educational organizations. We have specific programs for Libraries, Universities, and K-12 Schools. The ESRI Conservation Program oversees support, training and grants to thousands of conservation and indigenous organizations worldwide. It also set up conservation grant partnerships with other vendors and institutions that grant millions of dollars worth of equipment.
Finally and most importantly, ESRI is our users. They come from all ages, nations and walks of life. They typically share a belief that these GIS tools can help make the world a better place to live. They are often able to forge partnerships and relationships with their colleagues in conflicting institutions that canít be forged at higher levels. In so doing, they build bridges of information sharing and exchange that can and often do go a long way towards overcoming insitutional antagonisms. If you want to see what this looks like in action, come to the annual ESRI Userís Conference, the largest GIS gathering in the world.
Now that weíve seen the parts of ESRI, Iíd like to tell you a story about science and mapping that illustrates a lot of what I think is important about GIS. This is the story of Cassiniís Planisphere.
Giovanni Cassini was chief engineer to the pope in 1669 when Louis XIV, the sun king, appointed him director of the new Paris Observatory at the just-finished French Academy of Sciences. Cassini was already a famous astronomer, (known even today for Cassiniís division in the rings of Saturn), who had worked out a more accurate way of keeping time by basing it on the movements of Jupiterís moons. Time, of course, is the basis of longitude, which is why units of both are divided into 60 minutes and 60 seconds in a scheme dating back to the Babylonians. Until that time, maps were based on general descriptions by explorers and crude estimates of latitude and longitude, commonly with errors of hundreds of miles. Cassiniís Jupiter tables were accurate to within a minute, and in 1679 he began to establish a system of precise control points across France based on Jupiter moon time, accurate even by modern standards, and from these control points set out to re-survey the country. His survey moved the entire coastline, 200 miles from Paris by the best prior maps, a full 60 miles further inland. This caused his king to lament that his support to the sciences, intended to expand French power, had only served to diminish his kingdom! (Morrison & Morrison 1987).
Cassiniís success led him to begin a similar effort for the whole world. As his surveyors set out to establish precise standardized Jupiter control points in every corner of the world, he laid out an enourmous circular chart, called the Planisphere, on the limestone floor of the Observatoryís octagonal west tower. Over the next several decades, astronomers and explorers from all over Europe reported on their surveys and travels, recording them in the giant circular grid painted on the floor, and noting those points which had been precisely surveyed from Cassiniís own methods and standard controls. In so doing, by 1689 Cassini had created the first world map to be based on a framework of precisely-located standard reference points. This echoed the work of Hipparchus in defining the first standard system of spherical measurement, but Cassini was the first to use the tools of science to create a framework of standard control points to which all other positions could be compared and calibrated. For the first time, scientists had an accurate picture of what the known world really looked like, in a precise context that they could fit new observations into and slowly build upon their picture. His work led also to the first triangulation survey of France, which sparked the modern survey of England when vast errors were found as the French survey reached across the English Channel (Morrison & Morrison 1987), and established the non-spherical shape of the world, allowing Newton to prove that it was Earthís gravity that governed the motion of the Moon.
Cassiniís planisphere provided a precise framework for observations. A framework is an important idea in this regard. A framework is a skeleton or scaffolding into which a tool or a piece of data can be put that gives it meaning, context and a larger structure. Most scientific disciplines are frameworks, and in the study of science history they are also called paradigms, after the way they guide and focus the thinking of their adherents. Our use of frameworks is more general, including relational data models and software architectures.
The figure shows one example of what a framework can mean to a body of data. By itself, the data may be able to answer certain specific questions, depending upon the scope of the research and the original hypothesis, if any. Other questions about the significance of the results, or the effect of unrecognized assumptions, cannot be addressed without a broader context. A larger body of theory, or a data model representative of that theory, can explain relationships and mechanisms which help solve the broader questions. Basically, a framework provides an organizing principle that can gather and direct many individual efforts towards a common goal, such as biodiversity conservation. A framework can be an administrative coordination program like the national GAP program, a body of formal theory, a set of data model standards or software component standards, or a system of beliefs such as deep ecology. GIS is a framework which has expressions in all of these areas.
A very important concept with frameworks or paradigms is how they become established over time and how they change or remain rigid. A framework functions best when it organizes efforts around common outside goals. It functions poorly when it loses outside goals and instead focuses on itís own continued maintenance as the main objective. When outside goals change or new conflicting data is discovered this sort of framework canít adapt, and it is especially poor at comprehending ideas and concepts from other disciplines or outside points of view. Can we imagine a framework that is effective at interdisciplinary communication and constantly accountable to outside conservation goals even as technologies and data change?
| Software - Data - Books - Consulting - GIS Store | News - Events - Careers - About ESRI |
| Support - GIS Education & Training - GIS for Your Specialty - Partner Solutions - Library |
Copyright © Environmental Systems Research Institute, Inc.
Send your comments to: firstname.lastname@example.org
July 18, 2000