MSc Geographical Information Science 2010
Liam O'Connor

Why here and not there?
A GIS approach to graffiti


Graffiti is art, graffiti is crime, graffiti is Banksy, graffiti is an ASBO. Regardless of where you sit in any debate on the value of graffiti, there is always one indelible truth: it exists. And to exist it must exist somewhere. It is in the study of this 'where' that we hope to learn something about the essence of these scribbles on walls. We ask: why is graffiti in this place and not in that place?

In the end we find graffiti to be a duel construct, made up of the desires of a global subculture and how the city allows these desires to be made manifest.


In this project we collect, digitise, analyse and visualise graffiti in the context of its spatial distribution. Collection was achieved using a GPS enabled camera. A total of 802 graffiti points were collected within Edinburgh city centre data zone boundaries.

The analysis of the data was conducted by using established crime analysis techniques for the locating of hotspots of graffiti activity. The software package CrimeStat III provided most of the applications necessary for this task.

Subsequently ArcMap was used to visualise the results.

The project looked at many aspects of graffiti some of which are presented here.

Total sampled area set within greater Edinburgh in which all points of crew graffiti were collected & A Kernal Density estimation showing the distribution of hotspots in the city centre



We find that the physical factors which define the location of graffiti in the city centre are, for the most part, limited to perfunctory city constructs such as crossing guards, phone boxes and electricity boxes but that every now and again, city features come together in such a way as to create the opportunity for clusters of graffiti to exist. These hotspots occur in back alleys and yards, close to but not directly on the road network and generally consist of 'throw-up' type graffiti. These are slightly more elaborate efforts which take more time to complete and therefore occur in spaces just outside the constant view of potential distracters.
A second type of hotspot can occur out in the open street. This hotspot is the result of proximity to a legal graffiti space and the presence of a multitude of doorways which act like nooks in an otherwise highly visible space.

Showing a hotspot of the type found in yards and back alleys. The bulk of the hotspot is surrounded by the walls of a building, keeping it secluded and therefore this hotspot consists of the graffiti type 'throw-up'.

Showing a hotspot in close proximity to the legal graffiti wall. The bilk of the hotspot lies in doorways spread around the edges of this major junction.

Crew Territory

Subsequent to hotspot analysis, the project aimed to establish if different graffiti crews where more active in different parts of the city centre. Kernel Density Estimations were used to this end. In the map below we can see the four main city centre crews each exhibit different spatial patterns, this suggesting territoriality. OE have the largest central density, KSM are dominant in the south west of the city centre, youts north west, and NSA seem to have a large spread with slight dominance in the south east.

Two further areas outside the city centre were sampled to act as comparisons for findings in the city centre. The graphs presented below show again that territoriality may exist. With crews such as YLT having a large presence in Leith and none in the city centre, we again conclude that territory is a factor in graffiti. However, the results do not tell us why this territoriality exists.

rthrherthwghjgfhgjghjgfhergd A KDE showing the relative distribution of crews in the city centre ljnfwknf;kwmnv;a;knvlnrvwegaaergefvrvervevwdgjdfdfjnvevbsgbwq3 graphs showing the crews active in each area


To assess how the desires of graffiti writers have an effect on graffiti locality, the project looked at the idea of risk verses reward. The concept of 'fame' is well documented as a driving force behind graffiti. The best amongst his/her peers is the one afford the most respect for either 'getting up' more than anybody else or in getting your work seen by the more people. To this end the project looked at one particular instance of graffiti: that of the type 'throw-up' which occurs on a high wall. Viewshed analysis was used to see if this graffiti point was within the view of a close by CCTV camera.

A viewshed (yellow) cast from the CCTV camera (red point) towards a high wall graffiti point (blue).

This anlysis shows that in this instance the graffiti writer was willing to take the risk of being caught on camera in order to 'get up' on a wall of fame.


From analysis of surfaces and hotspots we find that the spatial distribution of graffiti is intrinsically linked to the positioning of city centre features and spaces but more specifically it is linked to the street network. From analysis of the different crews we see that territoriality may exist, however this is perhaps a territoriality of convenience rather than an idea that the crews violently mark out boundaries for their 'turf'. In the end we find graffiti to be a duel construct, made up of the desires of a global subculture and how the city allows these desires to be made manifest. These desires include fame amongst peers and the idea that reward outweighs risk.


  • Chainey, S. & Ratcliffe, J., 2005. GIS and Crime Mapping, Chichester, Wiley.

  • Ferrell, J., Hayward, K., Young, J., 2008. Cultural Criminology, London, Sage.

  • Halsey, M., Young, A., 2006. ‘Our desires are ungovernable’: Writing graffiti in urban space. Theoretical Criminology, 10(3), pp 275-306.

  • McDonald, N., 2001. The Graffiti Subculture. Youth, Masculinity and Identity in London and New York, New York, Palgrave Macmillan.
For further information on the research outlined in this paper see: O'Connor, L.,2010a; Why here and not there? A GIS approach to graffiti. Edinburgh, University of Edinburgh.