Introduction to Geographic Information Systems
Mankind has used maps for centuries, to mark territorial boundaries, trace the paths of migrating herds, and draw up strategic battle plans. Etched on cave walls, papyrus leaves, and finally paper, these maps not only provide a visual recording of man’s history and environment but their graphical nature transcends the boundaries of language and culture.
The traditional function of a map is to act simply as a guide, an aid to plot a path from one location to the next, however by incorporating additional geographically related information maps can be utilised to portray much more than simple geography.
Through the use of overlays (traditionally transparent sheets marked with pertinent data), background maps can be used as a reference to geographically analyse complex data, such as the spread of disease, environmental influences on pollution or troop movements.
The advent of computing, and in particular Computer-Aided Design (CAD), eased the creation of maps as cartographic data could for the first time be stored, edited and amended at will, without requiring a new chart to be scribed each time a change was made. However such systems provided only limited analytical functions, as their chief purpose was the capture and input of cartographic data.
Geographic Information Systems (GIS) however offer much more in the way of functionality than that of simple digital cartography. GIS provide three distinct functions;
data capture, manipulation and modelling.
Mapping information with a GIS can help to identify emergent patterns or relationships in geographically referenced data as it provides a dynamic perspective where relevant factors can be visualised and assessed in their effects upon the analytical model. Locational information stored in computer files or a database;
such as addresses, post codes, or cartographic co-ordinates, can be mapped to real-world business or ecological situations to provide a detailed analysis and unique visualisation not achievable by any other means. The purposes for such analysis vary from the assessment of environmental impact in council planning permissions, to the examination of customer demographics and location in order to choose a new site for a supermarket.
Currently many government and business organisations are converting, or looking to migrate, their paper-based maps to computer-based digitally-stored systems. The reasons for this are many and varied, though one reason is that digitally-stored cartographic databases can be altered much quicker and more easily than their conventional counterparts, and by locating the database in a central repository shared data can be standardised.
--------------------------------------------------------------------------------
Origins of GIS
The key element of a Geographic Information System is the separation of differing datasets into thematic layers, which can be overlaid upon a base map in order to visualise geographic relationships, or patterns. However this method of overlaying maps is neither new nor confined to advanced computerised GIS applications, and there are several historic examples of overlaying maps in this fashion.
During the American Civil War, French cartographer Louis-Alexandre Berthier, drew maps of the Battle of Yorktown which made use of hinged overlays showing troop movements.
The ‘Atlas to accompany the second report of the Irish Railway Commissioners’, produced in the mid-19th century, showed population, traffic flow, geology, and topography all displayed on the same map.
The 1854 London cholera epidemic was traced to a contaminated well by Dr. John Snow, by plotting the deaths of victims on a map.
Several key events led to the inception of GIS and digital spatial analysis, namely:
Improvements in computer technology, particularly graphic-handling capabilities, processing power and programming languages made such large-scale developments possible.
Development of theories of spatial processes, such as Freeman chain encoding, and in particular the research work carried out at the University of Washington Department of Geography by Nystuen (fundamental spatial concepts), Tobler (digital cartography and map projection algorithms), Bunge (geographical geometrics) and Berry (geographical matrix by characteristics) from 1958 onwards.
Increasing awareness of the environment and the potential benefits of GIS.
Early computers possessed no way of visualising spatial data using as their graphics handling was rudimentary to say the least and any systems capable of displaying anything other than character-based screens were extremely expensive. Because of the high cost of the required hardware, the first generation implementations of geographic information systems were developed and utilised by universities and government departments, particularly in Canada and the USA.
The world’s first large-scale operational GIS was Environment Canada’s Lands Directorate Canadian Geographic Information System, generally known as CGIS, funded by the Canadian government. The system was developed by a geographer named Roger Tomlinson in collaboration with IBM Canada as a direct result of the Agricultural Rehabilitation and Development Act of 1961. The main purpose of CGIS was the analysis of data collected by the Canada Land Inventory and the production of statistical summaries classifying the environmental properties of rural areas and parks, such as soil capability for agriculture and present use, in order to produce land management plans. The ability to produce simple maps was added as a later enhancement.
Although started in 1966 the database was not completed until 1974 as the project was hampered by high development costs. The lack of previous benchmarks and required technology forced the production of many innovative ideas, which subsequently became core functions of GIS, such as the use of scanning for the input of high density areas (an experimental scanner was built purposely for map input), vector conversion of scanned images, geographical partitioning (tiling) of large map areas, and the separation of data into thematic layers as well as the key concept of separating spatial and attributal data.
One of the key developments in digital cartography was Howard Fisher’s SYMAP (SYnagraphic MAPping), originating from the University of Chicago in 1964. SYMAP’s main use as a general-purpose mapping system was the handling of geographic data for display and output to a line-printer. Although the system possessed limited functionality and analytical operations it was simple to use and provided the first demonstration of the use of computers to generate maps. Fisher went on to set up the Harvard Laboratory for Computer Graphics and Spatial Analysis where much of the pioneering GIS work was undertaken.
Due to need of assigning census returns to their correct geographical location the US Bureau of the Census produced the world’s first geocoded census in 1970. As well as releasing traditional printed reports of the findings, the data was also released in digital form (using micro-data tapes), for users with access to computer facilities. The Dual Independent Map Encoding (DIME) files used by the Bureau contained far more detail than the printed reports, and where widely distributed and used as the basis for numerous applications.
The first commercially successful GIS was ArcInfo, developed by ESRI (Environmental Systems Research Institute) in the early eighties. ESRI was founded by Jack Dangermond, an early graduate of Fisher’s Harvard Computer Graphics Laboratory, and ARC/INFO was based upon many of the ideas and techniques taught there. The system was the first platform-independent GIS to run on mini-super computers, which made it an affordable prospect for many land resource management agencies.
The earliest UK example of a geographic database management system was LAMIS (Local Authorities Management Information System), a project based in Leeds and jointly funded by the Department of Trade and Industry, Leeds City Council and ICL. LAMIS was implemented in 1974 as an extended database management system with the capability of handling spatial information. The system was originally created to provide solidarity between files held by the various council departments. Integration was achieved by developing a core relational database, with each property given a Unique Property Reference Number (UPRN), this allowed the disparate Social Services, Rates, and Housing databases to be linked by a single common field.
Although the system is over 20 years old it is still in limited use, however the LAMIS database is no longer under development and it is soon to be replaced by a combination of ESRI’s ArcInfo and ArcView
(http://www2.prestel.co.uk/delphi/main.htm)
