Computational Approaches for Urban Environments: An Editorial - Computational Approaches for Urban Environments - page 3

and demographical aspects has been piled up promoting and supporting our
understanding of the operation of cities (e.g., Batty 2008 , 2013 ; Bettencourt 2013 ;
Glaeser 2011 ;Clarke 2014 ), the underlying processes and the interactions between
components are only partly understood thus far. Several academic disciplines are
dedicated to urban research. Each of them has their own unique perspective on
cities and utilizes their unique set of methods and tools. This results in fragmented
knowledge and a lack of coherent insights (Solecki et al. 2013 ). While qualitative
urban research has added significantly to the understanding of cities, past empirical
quantitative research has benefited from advances in the field of geographic
information systems (GIS; Goodchild 2010 ) technologies. At present, GIS-based
analyses (e.g., Jokar Arsanjani et al. 2014 ) have reached some level of maturity and
are an integral part of spatial sciences as well as of urban policy- and decision-
making. Although the quantitative analysis of urban areas is not new and goes back
to the quantitative revolution in geography (see Kwan and Schwanen 2009 ), the
rapid methodological progress - including spatial statistics, remote sensing, data
mining, and simulation-based modeling, among others - coupled with the recent
accumulation of readily available spatial and spatiotemporal data on a detailed
scale, i.e., volunteered geographic information (e.g., Jokar Arsanjani et al. 2013a ),
airborne laser scanning data (e.g., Xu et al. 2014 ), and cell phone data (e.g.,
Calabresea et al. 2013 ), among others, has stimulated and shifted the emphasis to a
computationally oriented urban science (Batty 2013 ).
In the literature, this linkage between geography and computational science
(Lazer et al. 2009 ) is referred to as geocomputation, coined by Openshaw and
Abrahart ( 2000 ). While the prefix “geo” emphasizes that geocomputation deals
with spatial theories, georeferenced data, and spatially explicit research problems,
the latter term “computation” highlights how geographical science is conducted,
namely, through a broad spectrum of computer-intensive methods, mathematical
and spatial statistical models, simulations, etc. Thus, geocomputation aims to
explore, extract, and generalize inherent urban patterns and processes, in data-driven
fashion from spatial and spatiotemporal data to not only solve complex geographical
urban problems but also to transform the implicit and hidden information in
spatial databases into urban knowledge. As such, geocomputation is an umbrella
term that includes, but is not at all limited to, agent-based modeling (e.g., Jokar
Arsanjani et al. 2013b ; Torrens 2012 ; Malleson et al. 2013 ), cellular automata
(e.g., Vaz et al. 2012 ; Pijanowski et al. 2014 ), spatial (e.g., Helbich and Leitner
2012 ) and spatiotemporal cluster detection (e.g., Nakaya and Yano 2010 ; Hagenauer
and Helbich 2013a , b ), Bayesian models (e.g., Brunauer et al. 2013 ;Lawand
Quick 2013 ), fuzzy logic (e.g., Grekousis et al. 2013 ), local regression modelling
(e.g., Leitner and Helbich 2011 ; Helbich et al. 2014 ), regionalization (e.g., Wang
et al. 2012 ; Helbich et al. 2013c ), and neurocomputing coupled with or without
evolutionary algorithms (e.g., Arribas-Bel et al. 2011 ;Guetal. 2011 ; Hagenauer
et al. 2011 ; Helbich et al 2013b ; Mimis et al. 2013 ). For a more detailed and thought-
provoking theoretical discussion, the reader is referred to Couclelis ( 1998 ), Fischer
( 2006 ), as well as Birkin ( 2009 ).