The Nature of Geographic Information Systems

The fact that a wide variety of data contains defined relationships to space means that a location in space can be used as a meeting ground for all of that data, regardless of its forms, content and assumptions. This is one of the most powerful concepts in GIS. Science and society are characterized by deep divisions between different disciplines, such as the social and physical sciences, and the data collected by these different disciplines reflects these divisions. There is no simple way to imagine combining demographic data collected about families with a soil health classification based on patterns of soil microflora populations, yet if those families were farmers it would be very important. Since both of these can be represented in space, families as street addresses and soil types as a classified landscape, they can be combined with a GIS to study issues such as the effect of poor soil biodiversity on family farm income.

An important concept in spatial integration is the spatial standard. GIS provides tools to make 2 different spatial data sources match each other, but without reference to a common basemap standard it is difficult to go any further. A spatial basemap provides a common framework that any data source can be registered to, and once registered, all other data meeting that same standard are immediately available for comparison with the new data. A basemap also commonly includes control points, precisely-located benchmark coordinates which allow the error and accuracy of positional data to be readily determined. The importance of controls in any scientific effort does not need to be reiterated.

Another important concept is scale. Basemaps of different scales represent completely different things in the landscape, mainly because each uses a different minimum mapping unit, representing the smallest landscape feature which can be captured and described in the database. Note that this concept of resolution applies equally to mapped locations and to feature descriptions and classifications. Most GIS projects will need to use several different scales, depending on the patterns and processes under study, and may therefore need to select several different basemap standards, and according, several different classification schemes. In the same way that basemaps can be nested in common sets of control points, classification schemes can be hierarchical, with finer divisions based on higher level categories.

One of the trends in computer development, especially recently, has been a move away from specific programming languages towards integrated software development environments offering a single platform with a selection of languages, interfaces and support tools. This works by basing the languages on a common architecture so that they can function compatibly with one another. From the very beginning, the goal of software development at ESRI was to produce tools that would function in whatever hardware and software environment a user wanted to utilize, unlike other GIS efforts based on single platform and proprietary designs. As a result, software development at ESRI has focussed upon building a well-engineered architecture for doing GIS functions within diverse and rapidly-changing technologies, and ESRI was a pioneer in the computer industry in offering a single software environment that functioned identically across multiple UNIX, PC and Mac platforms and integrated with every mainline database management system. This architecture has matured over 2 decades and now regularly incorporates new advances, such as ActiveX and dynamic web servers.

The figure shows some of the classes of operations commonly carried out in a GIS effort and illustrates the myriad functional relationships and interconnections between the different ESRI software modules and applications used to carry them out.

Because it is based on a common spatial data model and integrated many software tools within a single architecture, GIS permits the concurrent application of different theories on the structure of space and nature. This can be as simple as overlaying a demographic map with it’s implicit assumptions of the role of human motives in directing urbanization with a soil map and it’s implicit assumption of the role of soil fertility in determine land use. Failure to account for differences in implicit assumptions is a constant problem in data integration and data sharing.

Beyond this, GIS as an integrative framework makes it ideally suited for the study of ecology and geography, which themselves represent the union of numerous disciplines in trying to understand pattern and process in nature and in humans. The ability to explicitly manage spatial standards and basemaps provides important frameworks for the development of spatial models.

ESRI’s software also solved an old problem in geography and ecology, namely the differences between continuous patterns on the landscape and discontinuous ones. Continuous patterns like vegetation were best represented by continuous grid of cells, while discontinuous patterns like land tenure and point data were best represented by points, lines and polygons. When the GRID cell-modeling software was integrated into the polygon data model under ArcInfo several years ago, this problem finally became manageable. GIS integrates different spatial and attribute models using different methods for different kinds of theoretical analyses so as to create meaning and better decisions.

GIS is a central part of a number of planning and structured decision-making efforts, allowing the explicit management and recognition of different views of the landscape into a single managed process of consensus. It is also the most widely used computer technology among planning and land management agencies worldwide. It’s utility in conservation, however, is limited by political and social factors. The best available science must always be included in planning and land decisions, but decisions still usually go where the power is, and conservation can’t succeed unless there is the mandate and the political will to back up what the science proves. So an important precondition for integrating different players is to make sure they are all on the same field at the same level. Unfortunately, this fact of life is left out of most theories of computer supported collaboration and decision support.

Once there is some parity in political will, the integration of rigorous scientific data from different parties into a common framework can be extremely productive at illuminating precisely the areas of conflict, clarifying the effect of each individual contribution upon the whole, and suggesting the specific paths to resolution.

Before we finish, let’s take another look at Cassini’s Planisphere and what it represented as a framework for organizing information in response to human needs, compared to numerous similar efforts in biological data today. Cassini was trying to support Louis XIV’s desire to use the sciences in order to advance the French Empire worldwide. They needed one central location, accessible to the king where they could begin to form a view of what the world really looked like in order to better guide and advise French navigators. They weren’t especially interested in helping other countries navigate better. Nowadays we see this approach especially common in scientific databases, with center after center being built for the purpose of collecting and managing centralized scientific databases. What people are doing in these programs is solving problems for themselves or their institutions the way Cassini solved problems for France. This can be contrasted with another approach to technology and data, the service-oriented approach of solving problems for others. As you can see, an entirely different set of assumptions are needed in this case. Neither approach is "right" and in fact both are needed, but the problems occur when groups that succeeded in solving problems for themselves decide to shift to a services approach without a fundamental change in assumptions. That’s why we see programs offering high-end Unix stations as the "solution" for specimen data management. That’s why we see programs who purport to be making their data available to anyone who needs it but who don’t even have a data publishing program! It takes two distinctly different frames of mind to be successful in these two directions, and it is rare to find an organization that can successfully combine both.

Now let’s summarize by looking again at all these different things GIS is, all these different integrative frameworks. Do they have anything to offer biodiversity? Can they be combined into a single concept for the goal of biodiversity conservation?

This is a conceptual diagram of one possible way to arrange the capacities and activities in GIS with those of biodiversity institutions to promote the development and integration of better data and better tools for biology. It derives in part from already-functioning programs in technology transfer and already-established database efforts. The details of governance and administration are sketchy since as a GIS person I’m not as directly involved in those areas as you are, and I would hope that if such an idea finds merit, these details can be filled in by others. It may be that you already have plenty of such proposals and instead stumble on other problems I’m not aware of, but at least you can see here how we think the technology of GIS can have a contribution and perhaps you have in mind other productive ways to apply these ideas and our offer of help.