Networks of urban interaction - Growth and centrality in the complex geography of urban activity
Doctoral thesis, 2021

How cities and regions grow and decline depend on technological, social and economic factors. Understanding the interplay of these forces is central in research efforts aiming to improve urban and transport planning. The purpose of this thesis is to explore how mathematical modelling and computer simulation can contribute to these efforts and a central aim is to achieve practically useful models with retained conceptual simplicity as well as correspondence to important empirical patterns.

The approach combines a spatially fine-grained representation of land, with processes of urban interaction based on the theory of complex networks. Urban activity at a location is modelled as the sum of all economic interactions stemming from that location. The potentials for interactions and activity are deduced mainly from spatial constraints, such as transport networks and land use regulations. Concepts that are studied include urban growth, accessibility and urban agglomeration.

For model validation, an extensive data set on Swedish land taxation values is used. These values are based on actual sales prices and rent levels and can thus be considered as reasonable proxies for urban economic activity. Comparisons are made between empirical data and model outcomes, both with regard to probability distributions and geographical distributions.

The empirical probability distribution of land values is found to be well approximated by a power law, strengthening the case for modelling the system as a complex network based on a process of multiplicative growth. By combining these principles with spatial interaction mediated by a transport network, the preferential centrality model is formulated. The activity predictions generated by this model reproduces empirical geographical patterns of land values.

The presented models provide explanatory links between the structure of transportation networks and the geographical distributions of urban economic activity. This makes them attractive as starting points for the further step of creating practically useful planning applications. For example, the models could be used to assess how specific transport infrastructure improvements influence urban expansion.

urban activity

transport infrastructure

power law

urban growth

spatial interaction

accessibility

complex networks

Zipf’s law

centrality

KA-salen, Kemihuset, Kemigården 4, samt zoom: https://chalmers.zoom.us/j/65813664054, Password: 465103
Opponent: Marc Barthelemy, Institut de Physique Theorique (CEA), Saclay och Centre d'Analyse et de Mathematique Sociales (EHESS), Paris, Frankrike

Author

Alexander Hellervik

Chalmers, Space, Earth and Environment, Physical Resource Theory, Physical Resource Theory 2

The urban economy as a scale-free network.

Physical Review E,; Vol. 68(2003)p. 036124-

Journal article

A complex networks approach to urban growth.

Environment and Planning A,; Vol. 38(2006)p. 1941-1964

Journal article

Preferential centrality - a new measure unifying urban activity, attraction and accessibility

Environment and Planning B: Urban Analytics and City Science,; Vol. 46(2019)p. 1331-1346

Journal article

A. Hellervik, L. Nilsson, and C. Andersson. Preferential centrality as a multi-regional model for spatial interaction and urban agglomeration.

SAMMANFLÄTADE PLATSER, CENTRALA LÄGEN OCH VÄXANDE STÄDER

Ekonomiskt och socialt utbyte mellan människor sker i ett nätverk av kontakter som ständigt omformas och utvecklas. Det är lättare att skapa utbyte mellan närbelägna platser vilket gör att nätverket bäddas in i geografin. På så vis kan uppkomsten av städer beskrivas – platser flätas samman genom att kontaktnätverket tar sig fysisk skepnad i form av vägar, spår och byggnader.   

När tätorter och städer beskrivs som nätverk öppnas en möjlighet att hitta likheter med andra system i naturen och samhället. Det har nämligen visat sig att det finns många återkommande kännetecken hos stora och svåröverskådliga nätverk, så kallade komplexa nätverk. Med hjälp av datorsimulering och beräkningar är det möjligt att studera hur sådana nätverk växer, fungerar och utvecklas.

Gemensamt för många komplexa nätverk är: ”åt den som har skall varda givet”. Delar av nätverket som redan är väl uppkopplade drar till sig en majoritet av nya kopplingar, vilket skapar en ojämn statistisk fördelning. Sådana processer kan även förklara uppkomsten av skyhöga fastighetspriser i centrum av storstäder. 

Genom att studera stadens fysiska nätverk går det också att beräkna olika platsers centralitet. Det blir då möjligt att bedöma vilka lägen som har hög potential för täthet och intensiv verksamhet. Beräkningarna kan dessutom användas för att förstå hur nya vägar och järnvägar kan omforma städer. Detta gör metoderna direkt användbara för framtida planering av städer och transportsystem.

INTERACTION, GROWTH AND CENTRALITY IN CITIES AND URBAN LOCATIONS 

Economic and social interactions take place in a constantly changing network of contacts and trade. Connections are formed more easily between nearby locations, making the network embedded in the geographical context. In this way, the emergence of cities can be described – places become intertwined by the interaction network taking physical shape as roads, tracks and buildings.

To describe cities and regions as networks open up for finding similarities with other natural and artificial systems. Many recurring patterns can be observed in large and complex networks. Using computer simulations and mathematical modelling makes it possible to study the growth and function of such networks.

A common trait in many complex networks is that “rich get richer”.  This means that already well-connected parts of the network attract many new connections, creating an uneven statistical distribution.  Such processes can also explain extreme property prices in central urban locations.

Moreover, by analysing the physical networks in a city, it is possible to calculate the centrality of different locations, making it possible to identify the potential for high density and intensive activity. Such calculations can also be used to understand how adding new roads and rail links can influence the shape of cities, making the methods suitable for use in urban and transport planning.

Induced innovation cascades and the design of new sustainable processes for planning and policy: a modeling and empirical study of the development around road corridors

Formas (942-2015-124), 2015-09-01 -- 2018-12-31.

The transportation system as key to achieving societal aims through regional and urban development - empirical analysis, models and stakeholder involvement

Formas (2018-01038), 2019-01-01 -- 2022-12-31.

Driving Forces

Sustainable development

Areas of Advance

Transport

Subject Categories

Economic Geography

Other Natural Sciences not elsewhere specified

ISBN

978-91-7905-567-7

Doktorsavhandlingar vid Chalmers tekniska högskola. Ny serie: 5034

Publisher

Chalmers University of Technology

KA-salen, Kemihuset, Kemigården 4, samt zoom: https://chalmers.zoom.us/j/65813664054, Password: 465103

Online

Opponent: Marc Barthelemy, Institut de Physique Theorique (CEA), Saclay och Centre d'Analyse et de Mathematique Sociales (EHESS), Paris, Frankrike

More information

Latest update

10/7/2021