Geographic Information Systems
Basic and Advanced Functions of GIS
Originally, scholars used GIS for organizing data and creating maps for presentations. While still common, GIS software can now perform more analytical functions. The next two sections describe some of the descriptive (basic) and analytical (advanced) functions of GIS.
A. Descriptive GIS
Managing Data:
Managing the data for large projects is often difficult, and geographical information is especially cumbersome. GIS uses a framework for storing and accessing many forms of data, from tabular databases to images and maps. The software connects to these files and represents them in a traditional directory structure. Once the files have been imported to the project, they may be linked to spatial data and to each other. For instance, a researcher can create a historically accurate map of a county, link a census database to another database from a business directory, and then display various attributes of those people and businesses on the county map.
Creating and Overlaying Maps:
The end product of a GIS project is one or a series of maps – which are simply visual representations of data arranged in a spatial manner. These maps may display points of reference, analysis of phenomena, or elements of both. GIS facilitates basic cartographic tasks, such as creating legends, north arrows, and distance scales. It also performs the difficult job of accurately projecting a map to the shape of the Earth’s surface. A GIS also combines any number of layers and allows their transparency and sequence to be adjusted. Researchers can therefore combine and compare maps from different periods and maps that highlight different geographic features. As well, they can overlay representations of their own historical analysis on historical or modern maps.
B. Analytical GIS
Below are some of the principal ways scholars have used GIS to analyze spatial relationships in historical data.
Joining Data:
Creating thematic maps, or maps that display quantitative attributes, often requires joining data. Linking data to a map requires that the database(s) have a geographic attribute in common with the map, be it a coordinate, a town name, a lot number, or a street or concession address (see the Case Study). Censuses and business directories are well suited to this linkage because they are ordered according to geographic location. Linking two databases in a GIS requires that they have a common field. Once this field has been standardized the software can form a join based on the variable, and the data now found in one table may be linked to another, and, ultimately, a spatial representation of the information. For instance, a business account book that recorded sales of goods to customers in different towns may be compared to the demographic features of those towns and the distances between each town and the business. Or if the business recorded transactions with rural residents, and the accounts can be linked to census data by the client’s name or lot number, GIS offers a variety of analytical tools for understanding these rural relationships.
GIS allows integrating and relating data from a variety of sources. In the discipline of history, there are a wealth of paper maps and statistical information buried in cabinets, atlases, and census records that can be digitized, analyzed, and displayed in a GIS.
Georeferencing:
One problem with using today’s spatial information for historical study is that landscapes change over time and human geography changes even more quickly. Place names, populations, and land use change frequently, but topographical characteristics such as forest growth and soil quality also evolve. Georeferencing is a tool in GIS that allows researchers to overlay maps made from modern spatial data with more historically accurate maps, surveys, and air photographs (photographed for many places in the early twentieth century).
One form of georeferencing is known as “rubber sheeting,” where researchers identify several equivalent points, such as boundaries or intersections, on the modern and historical maps. The GIS software then reorients and projects the historical documents so they fit modern coordinates. Then, researchers may reliably compare data at any given point, or even display the results of their analysis overlaid on historical maps and images.
Spatial analysis:
Studying historical data is often facilitated by projecting those data over space. But, sometimes the geographical information about a place or its environment is one of the only ways to make sense of certain data. For example, the production of certain commodities is often based heavily on the producer’s proximity to natural resources or modes of transportation. With GIS, researchers can more easily measure the distance between points and analyze the ratios of production to distance.
The software takes many other environmental factors into consideration. For instance, elevation data can recreate the “viewshed” of a particular area, or the points visible to an observer at any given location. This has helped archaeologists understand the placement of human objects and landscaping and predict the location of other archaeological sites. Others use spatial analysis to find correlation between settlement or other human activity and geographic features such as altitude or drainage.
Geostatistical analysis, or Interpolation:
GIS can use geographical attributes to help “fill in the blanks” when information is available for some places but must be interpolated for others. This tool uses correlation and regression analysis and considers environmental variables when predicting the most likely characteristics of any unknown point in a spatial grid. Geostatistical analysis facilitates the modeling of regional economic development based on environmental factors and the distance between points. It has helped historians understand the effect of drainage on disease, and spatial interpolation has also been used to study migration where changing census boundaries have frustrated other quantitative historians. Researchers who use census samples may also experiment with interpolating the new data they retrieve from cross-tabulations of manuscript census databases.