2.2 Building a GIS Vocabulary
2.2.5 Representing the World III: Location
Determining Location - Coordinate Systems and Projections
Every feature in a GIS has a location, but how is that location defined? We discuss the location of things and use geographic information all of the time, whether we are describing where we live, the location of a restaurant, or how to get to a particular park in the city in which we live. In these sorts of conversations we rely on a lot of contextual information. Geospatial data, however, uses mathematical coordinates mapped to the surface of the Earth according to a specific geographic coordinate system. Let’s differentiate a few useful terms before moving on:
Geographic Data encodes location in human-understandable ways (place names, addresses, codes).
Geospatial Data encodes location with coordinates (geometric representations according to particular geographic coordinate systems).
A Geographic Coordinate System uses a three-dimensional spherical surface to define locations on the earth with latitude and longitude.
You’ll notice the implication that there is more than one geographic coordinate system. In fact, there are two major types of coordinate systems (only one is a geographic coordinate system; the other is a projected coordinate system) and a number of different models for each of those two types of systems:
Geographic coordinate systems are good for representing where something is on a globe, but most maps are flat. To visualize geospatial data on a flat map, we use projections that transform the 3D Earth into 2D space. This is no easy task and no matter how it is done, there will be distortions. Imagine, for example, peeling an orange and trying to lay the peel flat on a table. You’ll have to stretch, tear, or squash it. This is the problem that map projections are trying to solve: taking the curved surface of the Earth and flattening it onto a 2D surface like a screen or piece of paper. Every projection makes trade-offs.
When you’re working at regional or local scales (like mapping a city, analyzing land use on a campus, or tracing travel in a specific historical region) projected systems, especially those optimized for a particular region, offer better accuracy than geographic coordinate systems. For example, UTM (Universal Transverse Mercator) Zone 33N is a common choice for much of central Europe, while British National Grid is preferred for work in the UK.
Choosing the wrong projection for your area or purpose can lead to distorted distances, misaligned features, or inaccurate measurements. Fortunately, tools like QGIS and ArcGIS allow you to reproject your layers so that everything lines up, as long as you know what coordinate systems you’re working with.
Coordinates vs. Place Names – Direct and Indirect Location
We already introduced one important distinction in specifying and recording location when we discussed geographic vs geospatial data, but let us add another: direct vs indirect georeferencing. Direct georeferencing uses precise coordinates (latitude and longitude or projected x/y values) to define a location.
For example, we might directly georeference the Eiffel Tower with its latitude and longitude = 48.8584 N, 2.2945 E.
Indirect forms of georeferencing objects are especially common (and of vital importance) in humanities research. Think of literary texts, archival documents, oral histories, or historical records: they’re full of locations, but rarely given in coordinates. Instead, they use narrative, cultural, or relational descriptions. To use this kind of data in GIS, you must geocode it: convert text-based locations into usable coordinates. This process is rarely straightforward. As Karen Kemp writes,
“in order to use a geographic reference that is not a coordinate, it is necessary to have or to construct the geographic framework of the named locations. Gazetteers are one major source of this kind of framework, but there are others” (Kemp, 2010, p. 45).
Gazetteers (structured databases of named places and their locations) are essential tools for this kind of work. But even with them, ambiguity remains. Place names may change over time, refer to multiple locations, or lack clear boundaries. Constructing a geographic framework for historical references can be extremely challenging, particularly when working with literary texts or archival materials. Kemp points to Roth Mostern’s work as an example of the rigor involved, noting that
“Mostern, a historian who has written extensively on the design of gazetteers as a foundation for historical GIS, has spent several years mapping the changing geography of administrative units during the Song dynasty in China from geographic descriptions recorded in administrative documents from that period.”
Her work highlights just how much interpretation, historical knowledge, and manual labor go into building a usable spatial dataset from indirect textual references (Mostern, 2008). In humanities research, as one might imagine, indirect georeferencing is often the default, not the exception. Recognizing the uncertainties, interpretive decisions, and contextual dependencies involved in translating names to coordinates is part of doing responsible spatial scholarship.
Why It Matters
The same set of geographic coordinates could point to two different locations depending on whether your mapping software is using WGS 84 or the Australian Geodetic Datum 1984. Knowing which one is being used can be a matter of life-or-death. Imagine, for example, a hiker is in trouble and sends a message for help with a set of geographic coordinates. If those coordinates were derived using a system that relies on WGS 84 trying to find the hiker using an application that relies on a different coordinate system—Australian Geodetic Datum 1984, for example—could point you to a different location. For more information on coordinate systems, see Heather Smith’s discussion on the ArcGIS Blog, maintained by ESRI, at https://www.esri.com/arcgis-blog/products/arcgis-pro/mapping/gcs_vs_pcs.
- Smith, H. (February 27, 2020) “Geographic vs Projected Coordinate Systems,” ArcGIS Blog. https://www.esri.com/arcgis-blog/products/arcgis-pro/mapping/gcs_vs_pcs
- Kemp, K. (2010) “Geographic Information Systems and Spatial Analysis for the Humanities”. In David Bodenhamer, John Corrigan, and Trevor Harris (ed.), The Spatial Humanities: GIS and the Future of Humanities Scholarship. Indiana University Press (31-57)
- Mostern, R. (2008). “Historical Gazetteers: An Experiential Perspective, with Examples from Chinese History,” Historical Methods: A Journal of Quantitative and Interdisciplinary History 41 (1), 39-46. https://doi.org/10.3200/HMTS.41.1.39-64.