I was digging through some material from my MSc in Geocomputation from Maynooth University and came across this exam question. The MSc was fantastic. We had access to the brains of Chris Brunsdon the late Martin Charlton, two of the three people (Stewart Fotheringham, another stalwart in geospatial statistics was the third) that gave us Geographically Weighted Regression. Anyway, you can read my answer from 2015 below, haven’t update or changed a word.
Defining GIS and GISci
First, let’s set a definition for both Geographical Information Systems (GIS) and Geographical Information Science (GISci) and then delve further into each definition to highlight the differences between them.
GIS
A GIS is primarily a computer system (hardware and software networks) capable of capturing, storing, manipulating, analysing, and displaying geographically referenced information; that is, data identified as having a spatial component/location. Practitioners also define a GIS as including the procedures, operating personnel, and spatial data that goes into the system (Foley, 2014 and Wright, n.d). GIS was born in the 1960’s with Roger Tomlinson’s Canadian Geographical Information System (CGIS) acknowledged as the first implementation (Kemp, 2008).
GISci
The term Geographic Information Science was coined in the early 1990’s by Michael Goodchild (Kemp, 2008 and Wright, n.d.). GISci aims to address the fundamental issues underlying GIS, mainly in the use of digital technology to handle and represent geographic information, and their use to advance scientific understanding (Kemp, 2008). GI Scientists study the (mainly digital) representation of the world and not the world in its concrete form. This is something that distinctly differs GISci from other sciences (Blaschke & Eisank, 2012). GISci encompasses the existing technologies and research areas of GIS, cartography, geodesy, surveying/geomatics, photogrammetry, digital image processing (rasters), remote sensing, and quantitative spatial analysis and modelling, and as such is the sciences behind the GIS application to accurately represent the Earth (UCGIS, n.d and Wright, n.d.).
Discussing GISSci
These definitions set GIS and GISci apart, that GIS is the end-user technology made possible by advancements in GISci. GISci, however, incorporates and requires knowledge and principles from other scientific disciplines and exists in symbiosis with these as it cannot exist without them (Blaschke & Eisank, 2012). GISci research includes, but is not limited to, spatial data structures, spatial data analysis, accuracy in representations, cognition, and visualization. These overlap GISci with several other traditional disciplines that are concerned with the Earth’s physical process and of how humans interact with the Earth such as geography, geology, geophysics, oceanography, environmental sciences, and spatial statistics analysis. Overlapping disciplines, more outside this list, include how humans interact with machines such as computer, information, and cognitive sciences, and artificial intelligence (AI). It is important to note that GISci, however, is not central to any of these disciplines, and represents a collaborative scientific approach that is defined by researchers from diverse backgrounds, working together on interrelated problems. These problems are scientific in nature but also service the needs and requirements of several government, industry and business related interests (Wright, n.d.).
GIScience revolves around short-term and long-term research agendas Short-term research priorities include GIS and Decision Making, Geocomputation, Geospatial Semantic Web and Information Semantics, Dynamic Modelling, and Temporal GIS. Some long-term research challenges include Scale, Geographic Information Engineering: Distributed and Mobile Computing, Space and Space/Time Analysis Modelling, and Geovisualization. GIScience as a field of study has now reached a mature level. since its inception in the 1990’s, with its own core curriculums, degree and certificate programs, scholarly journals, blogs, textbooks, research institutes and conferences (Wright, n.d.).
Discussing GIS
Modern GIS are a connection of hardware, software, databases, people as both operators and data collectors, and procedures all linked via computer networks. This networked system allows datasets from across the globe in diverse holdings to be combined together. GIS software is packaged to suit the requirements of assembling this data for its appropriate needs such as simply viewing the data through onscreen or printed output. The onscreen may be private on desktop software or shared on the web via a web mapping application. Another need might be to analyse spatial data to solve a problem or inform better decision making and as such, GIS are capable of advanced analysis functions (Kemp, 2008).
These advanced analysis functions are a part of the powerful suite of tools offered by GIS. GIS are a software product with a well-defined set of functions, but their efficient and effective use requires an understanding of numerous basic principles. One example is that when using a GIS the user should be aware of the utilization of scale when representing geographic data. It is currently impossible for a GIS or spatial database to contain all of the geographic detail found in the real world, although this is something that the vision of CyberGIS strives to diminish in the future (Goodchild, 2010). Therefore, all GIS employ some strategy with respect to scale (Kemp, 2008). It is important to note that GIS allow digital representations of the real world and as such these representations will be prone to different levels of inaccuracies at different scales. The Modifiable Areal Unit Problem (MAUP), identified by Gehlke & Biel in 1934, but was not brought to prominence until the 1970’s, is a scale issue involving ecological fallacies of aggregated data at different scales. This is a persistent issue for GISci, as mentioned earlier in long-term research, and something that the GIS user community should be aware of when aggregating data at various scales.
Someone trained in the use of GIS technology would be able to carry out routine operations but only an educated individual with the knowledge of underlying principles would be effective in devising new applications, troubleshooting problems, and have the ability to adjust to new and improved features of GIS technology. As such, suitably educated and qualified people are essential to the design, programming, and maintenance of GIS. These people also supply GIS with the spatial data and are responsible for analysing this data and interpreting the output (Kemp, 2008).
Waldo Tobler’s first law of geography states that “everything is related to everything else, but near things are more related than distant things”. This law is fundamental to the creation and interpretation of GIS representations and the concept can be formally measured as the property of spatial autocorrelation (along with temporal autocorrelation, “the past is the key to the present”). This makes a fundamental geographic statement (Kemp, 2008) and has allowed for the development of analysis techniques such as spatial interpolation (kriging for example).
Conclusion
In essence, Geographical Information Science revolves around research agendas for the advancement of Geographical Information Systems. GIS, as an applied problem solving technology, acts as a tool to address significant societal and environmental issues as well as conduct research for a multitude of individual and interrelated scientific disciplines. Therefore, GIS provides the environment in which the fundamental principles of GISci can be applied to current real world problems.
References
Blaschke, T., and Eisank, C., (2012) giscience.org, How Influential is Geographical Information Science? [online]
Available at http://www.giscience.org/past/2012/proceedings/abstracts/giscience2012_paper_104.pdf, (accessed on 28/04/2015)
Foley, R. (2014) GIS, NCG601A: Geographical Information Systems, Maynooth University, Ireland, Department of Geography, unpublished.
Gehlke, C.E., and Biel, K. (1934) Certain Effects of Grouping Upon the Size of the Correlation Coefficient in Census Tract Material, Journal of the American Statistical Association, Vol 29, No. 185: 169-70
Goodchild, M. F. (2010) Twenty years of progress: GIScience in 2010, Journal of Spatial
Information Science, Number 1, 3-20.
Kemp, K.K. (2008) Encyclopaedia of Geographical Information Science, California, USA: Sage Publications, Inc.
UCGIS, Organ State University (n.d.), Geographic Information Systems and Science [online]
Available at http://geo.oregonstate.edu/ucgis/ucgis_info.html, (accessed on 28/04/2014)
Wright, D.J., Oregan State University (n.d) GIScience_all.pdf [online]
Available at https://ir.library.oregonstate.edu/xmlui/bitstream/handle/1957/19044/GIScience_all.pdf?sequence=1, (accessed on 28/04/2015)