Cunningham maintained that communication is critical in interdisciplinary research because different disciplines often look at the world in different ways. The problems we face do not have instruction books so we must communicate effectively to solve them.
A Geographic Information System (GIS) is a group of elements or subsystems which are used to gain knowledge about the Earth, its features, and their distribution. The major components of a GIS are: hardware, software, digital data, procedures, and people. The major subsystems of a GIS are: data input, data storage and retrieval, data manipulation and analysis (that is what really sets GIS apart from other spatial information systems), and data reporting (output).
A key ability of a GIS is to take implicit information in a map and make it explicit (land form and slope from contours; distance to/from a feature). Cunningham quoted Ron Abler as saying that "GIS technology is to geographical analysis what the microscope, the telescope, and computers have been to other sciences."
Maps form the basis of any GIS. The basic elements of a map are scale, projection/coordinate systems, and symbolization. The fundamental problem of cartography is that there is a round Earth and a flat map. A map projection is the systematic representation of all or part of the surface of a round body on a plane. Because map projections represent a 3-D surface in only 2 dimensions, all map projections have some type of distortion. Map projections are usually based upon one of three types of developable surfaces: cylinder, cone or plane.
Many map projections exist to capture specific characteristics with the least distortion possible. No single projection can correct all errors; there is no "best" projection. Map projections generally address four types of distortion: area, shape, scale, and direction. The most common projections used are conformal; shape, angle and scale are generally correct, though scale is generally enlarged. Map projections are based upon datum. Horizontal datums are base on an ellipsoid and vertical datums are based on a geoid. Global Positioning Systems (GPS) is an Earth-centered coordinate system.
Cartographic primitives represent data on maps. They are either points, lines or polygons. These are discrete features. However, many natural phenomena are continuous. Discrete data is well represented by points, lines and polygons; these are called vector systems. Continuous data is best represented by a more regularly sampled array or surface; these are called raster systems. There are advantages and disadvantages of both which can be summed up in the quip: "Raster is faster, but vector seems more corrector."
GIS begins with building a digital spatial database. Creating this database is the most expensive aspect of any GIS project, taking from 50-90% of the total time invested. Methods of entering data include: manual encoding (typing in the data; slow, tedious, prone to error), manual digitization (tracing a map using a digitizing tablet), scanning or semi-automated processing (not a panacea, still requires lots of editing time), or direct digital transfer (transfer data from existing digital sources; faster, but still requires editing time, translating, checking).
Critical issues with regard to data input:
The ability to manipulate and analyze spatial data is what most clearly distinguishes GIS from other spatially referenced systems such as automated mapping and computer aided design. Beginning with a digital dataset, GIS provides the ability to interact with spatial data, to redefine, reassign and recombine spatial data, and to construct cartographic models, or a set of procedures to solve a problem. Three major types of spatial analysis are location, neighborhood, and zone.
Locations combine spatially discrete data, traditional overlay analysis, and local operators (any mathematical operation can be performed between layers or themes, for example, dividing the number of people by the number of households to get people per household). Neighborhood analyses analyze data within a prescribed distance. One type of neighborhood analysis is the proximity analysis which, for example, could identify all residences within five miles of a particular hazard. "Scan" or "focal" operations can determine minimum, maximum, density, average, diversity, etc., for example, determining the area with the greatest diversity of vegetation. Zone analyses analyze data within a prescribed boundary. Zonal operations might include summarizing conditions within a political unit, such as the percentage of agricultural land in a country.
Cartographic modeling is the combination of several layers of data through a prescribed set of procedures to create information. For example, it might be used to find the best place to site a landfill based on geology, hydrology, land acquisition costs, proximity to existing roadways and residential areas, etc. It is important to recognize however, that there are limitations to the technology. It does not always yield the right answer, and answers are only as good as the quality of the data going in. GIS is really two dimensional, so it has problems with 3-D and even more trouble dealing with time. There are also problems with how to deal with multiple scale and multiple resolution data.
Cunningham discussed several applications of GIS that he has worked on. One involved integrating GIS with an environmental simulation model for the Big Darby Creek Research Project in Ohio. The situation was that poor agricultural practices such as over-tilling had caused a great deal of soil erosion and siltation of Big Darby Creek. Using GIS to map the area revealed that over 90% of the land in the watershed is in active agriculture. The GIS also identified slope, soil information, 107 threatened and endangered species, and in creasing urbanization. Once baseline information was established, the GIS was used to predict sediment levels based on four scenarios: historic baseline, current conditions, adding a forest buffer strip along the stream to filter nitrogen and sediment, and a management policy that implemented the conservation reserve program. Four scenarios for nitrogen fertilizer application were also tested out in the GIS.
In another example, GIS was used in the service of the Cincinnati Hillside Trust, a project aimed at protecting the hillsides of Cincinnati. The GIS was used to simultaneously look at soil associations, land slopes, orientation, etc. The power of the GIS is in creating derivative data. It combines traditional land use maps with land cover characteristics from satellites, making data more valuable by combining it with other data. The GIS could also be used to study landslide patterns and provide predictions, by combining data on slope and geology and existing landslides; a landslide map could be created. Similarly, a GIS could create a model of ecological corridors or predict where development will occur based on the elements development generally follows. Combining several GIS models could predict where the most pristine areas are likely to be developed, with obvious policy implications. A policy matrix could be created based on this information.
In the development of GIS, there is a goal of technology transfer from government to private industry. One such example is the use of GPS in construction and mining operations. By placing a GPS on a bulldozer, the operator can be directed to avoid hazardous conditions and can properly position a building with respect to orientation. Tests have found such devices to be within 5 cm accuracy for real time positioning of x, y, and z coordinates. GPS is critical to GIS because it gives the spatial reference necessary.
Another allied technology is called "terrestrial photogrammetry" which records digital pictures from a van as it drives along. This technology reveals features of the land in great detail, collecting features for a county in one third the time usually needed to collect far less accurate data. Its greater accuracy and detail for less time and money is beginning to bring this technology into use in local mapping projects.