Urban models

- theoretical ideas about what cities are and how they function

- normative practices about how we should plan and build cities

- analytical representations to study and simulate the structure and evolution of cities

Urban model types

- theoretical ideas about what cities are and how they function

- normative practices about how we should plan and build cities

- analytical representations to study and simulate the structure and evolution of cities

empirical models

simulation models

abstract models

Urban modelling

Complementary epistemologies to study cities and urbanisation

- tackle causality in opposing ways

- tackle past, future and alternative timelines very differently

- tackle space, scales and contingency very differently

Challenges

- validation, integration, visualisation, theory-building, etc.

- demanding array of skills and knowledge domains

- unquestioned urban model classics vs. ad-hoc simulations

[Pumain and Reuillon, 2015]

Two strategies to improve urban models

Looking back:

A series of seminars on classical analytical urban models

With Rūta Ubareviciene & Ali Sobhani

Problem

Models are, by construction, simplifications of reality, but some simplifications have important consequences for urban studies and policy making, especially when their users are unaware of the original design choices of the models, which can happen when the urban model is very complex or has reached a “classical” status.  

Question

What are the unforeseen consequences of urban models’ simplification on our understanding of cities?

Organisation

11 speakers

8 models

4 topics addressed:

- what is the model doing?

- what is its biography?

- how did it influence urban studies?

- how did it influence policy?

90-minute online seminars

edited into 45-min podcasts

... coming article

Objectives:

> understand the context of modelling choices

> understand the effect of translation and transfer across disciplines.

Contexts of modelling choices

[Hegselmann, 2017] [Radeff, 2019] 

[Batty, 2022] [Sarkar, 2022]

Transfers and translations

[Hegselmann, 2017] [Ferretti, 2021] 

[Radeff, 2019] [Pumain, 2021] 

[Delloye, 2021] [Lerousseau, 2021] 

Looking forward:

A manifesto for geosimulation

With Denise Pumain, Paul Chapron, Romain Reuillon, Sébastien Rey-Coyrehourcq & Juste Raimbault

Problem

Generative simulation models have been "promising" for a long time now. They get even more "promising" with the development of big data and HPC. However, the process of building them is wasteful and suboptimal. The wheel gets reinvented instead of cumulative knowledge building and real theory development.

Questions

What are the obstacles to a cumulative development of geosimulation models?

What can be done about it?

Sources

Collective modelling experience in a multidisciplinary team tackling a similar question (2011-2015):

how to model urbanisation and the development of different urban systems over time?

Reflections and discussions from a recurring workshop on theory-building gathering agent-based modellers from various disciplinary fields (2017-2022)

What are the obstacles to a cumulative development of geosimulation models?

1. Awareness of existing models

> time, search, publication

2. Readability of the models

> language, architectures, entities, processes, etc.

3. Access to the models

> openness, archive, versions

4. Data

> access, usage, compability

5. Incompatibilities

> conceptual, technical, purpose, scales

[Cottineau, 2022]

What can be done about it?

A geosimulation meta-language

“We are going to go out on a limb and argue for a geosimulation metalanguage, which can present the model’s algorithms, as well as the characteristics important for model dynamics, as an exact sequence of model events, synchronous/asynchronous updating, criteria of self-organization involved, etc. [...] Currently, the cogs of the simulation infrastructure—if not the entire wheel—are being reinvented over and over by different groups, simultaneously. If converged to a common denominator, these attempts can give birth to such a language.”

[Benenson, Torrens, 2004, p.237]

4 conditions to a meta-language

1. Model reviews

> attentive to polysemy of concepts and diversity of disciplinary traditions tackling the same process

2. Version-tracking development

> crediting past contributions and focusing on evaluation/exploration

3. Modular infrastructure

> reusable building blocks. Assembly as theoretical labour

4. Painless integration

> of blocks, analysis methods, computing capabilities, viz, etc.

[Cottineau, 2022]

First examples

Create unexpected combination of rules

> automate the generation of model structures using genetic algorithms to combine building block/theoretical mechanisms.

Create a complete universe of mechanism interactions

> automate the calibration of model structures using genetic algorithms to analyse the hierarchy of explanatory power of building block/theoretical mechanisms.

[Cottineau et al., 2015]

[Vu et al., 2020]

Future of Cities

• Made of models, old and new

• Designed with models

• Discussed with implicit / conflicting models in minds

• Should include pluralism

• Should apply beyond contingent contexts

• Should build on all existing knowledge

• Batty M., 2022, "Forrester's Urban Dynamics & cities as systems", Urban Models Seminar, https://www.youtube.com/watch?v=yYPO4qFjJ_E

• Benenson, I., & Torrens, P. (2004). Geosimulation: Automata-based modeling of urban phenomena. John Wiley & Sons.

• Cottineau C., 2022, "Generative modelling" In. Harris R., Heppenstall A., Wolf L., A Research Agenda for Spatial Analysis, Elgar.

• Cottineau, C., Reuillon, R., Chapron, P., Rey-Coyrehourcq, S., & Pumain, D. (2015). A modular modelling framework for hypotheses testing in the simulation of urbanisation. Systems, 3(4), 348-377.

• Delloye J., 2021, "Alonso's model and urban centrality", Urban Models Seminar, https://www.youtube.com/watch?v=Pfyuv9pf6Wo

• Ferretti, F. (2021). History and philosophy of geography II: Rediscovering individuals, fostering interdisciplinarity and renegotiating the ‘margins’. Progress in Human Geography, 45(4), 890-901.

• Hegselmann, R. (2017). Thomas C. Schelling and James M. Sakoda: The intellectual, technical, and social history of a model. Journal of Artificial Societies and Social Simulation, 20(3).

• Heppenstall, A. J., Crooks, A. T., See, L. M., & Batty, M. (Eds.). (2011). Agent-based models of geographical systems. Springer Science & Business Media.

• Lerousseau J., 2021, "Economic Base Theory and urban development", Urban Models Seminar, https://www.youtube.com/watch?v=E8FSJmlALpM

• Lorscheid, I., Berger, U., Grimm, V., & Meyer, M. (2019). From cases to general principles: A call for theory development through agent-based modeling. Ecological Modelling, 393, 153-156

• Pumain D., 2021, "Central Place Theory and urban hierarchies", Urban Models Seminar, https://www.youtube.com/watch?v=YdB4e39EZEc

• Pumain, D., & Reuillon, R. (2017). Urban dynamics and simulation models. Cham: Springer International Publishing

• Radeff, A. (2019). Walter Christaller traduit en chinois (1998-2010). Cybergeo: European Journal of Geography.

• Sarkar S., 2022, "Utilitarian models & urban resource allocation", Urban Model Seminar, https://www.youtube.com/watch?v=5eUJyxLwnk0

• Vu, T.M., Buckley, C., Bai, H., Nielsen, A., Probst, C., Brennan, A., Shuper, P., Strong, M. and Purshouse, R.C., 2020. Multiobjective genetic programming can improve the explanatory capabilities of mechanism-based models of social systems. Complexity, 2020.

An evolution of analytical urban models and geosimulation

Clémentine Cottineau, BK TU Delft, Urban Studies

c.cottineau@tudelft.nl

@ClementineCttn

Citizens Collective Seminar Series

22 June 22