Openly accessible

Towards automatic calibration of neighbourhood influence in cellular automata land-use models

Roodposhti, Majid Shadman, Hewitt, Richard J. and Bryan, Brett A. 2020, Towards automatic calibration of neighbourhood influence in cellular automata land-use models, Computers, environment and urban systems, vol. 79, doi: 10.1016/j.compenvurbsys.2019.101416.

Attached Files
Name Description MIMEType Size Downloads

Title Towards automatic calibration of neighbourhood influence in cellular automata land-use models
Author(s) Roodposhti, Majid Shadman
Hewitt, Richard J.
Bryan, Brett A.ORCID iD for Bryan, Brett A. orcid.org/0000-0003-4834-5641
Journal name Computers, environment and urban systems
Volume number 79
Article ID 101416
Total pages 13
Publisher Elsevier
Place of publication Amsterdam, The Netherlands
Publication date 2020-01
ISSN 0198-9715
Summary © 2019 The Authors Cellular Automata (CA) land-use models are widely used for understanding processes of land-use change. However, calibration of these models is a knowledge-intensive and time-consuming process. Although calibration of common driving factors such as accessibility (A), or suitability (S) is a relatively straightforward task, calibrating the neighbourhood dynamics (N), which controls the key model behaviour, is often very challenging. Here, building on the SIMLANDER modelling framework, we develop an automatic rule detection (ARD) procedure to automate the calibration of N. To demonstrate the performance of the tool, we simulated 15 years of urban growth in Ahvaz, Iran (2000–2015) using a wide range of different rule-sets. We evaluated calibration goodness-of-fit for each rule-set against a reference map by means of cross-comparison of multiple metrics using a ranking procedure. The ARD procedure can be implemented in two ways: 1) by random sampling of the parameter space, a user-defined number of times, or 2) through a stepwise “grid search” approach for a user-defined number of rule combinations. Both approaches were found to produce successful rule combinations according to the goodness-of-fit metrics applied. Grid search performed slightly better, but at the cost of a fivefold increase in computation time. The ARD approach facilitates model calibration by allowing rapid identification of the optimum ruleset from a wide range of possible parameter settings, while the ranking procedure facilitates comparison of simulations using multiple metrics. The approach we present also helps to improve simulation accuracy with respect to manual calibration methods, and increases understanding of neighbourhood dynamics for the urban area studied. To encourage repeatability and transparency, our approach uses only open data and Free-and-Open Source Software (RStudio environment) and we provide our ARD scripts as supplementary material
Language eng
DOI 10.1016/j.compenvurbsys.2019.101416
Indigenous content off
Field of Research 0909 Geomatic Engineering
1205 Urban and Regional Planning
HERDC Research category C1 Refereed article in a scholarly journal
ERA Research output type C Journal article
Copyright notice ©2019, The Authors
Free to Read? Yes
Persistent URL http://hdl.handle.net/10536/DRO/DU:30130411

Connect to link resolver
 
Unless expressly stated otherwise, the copyright for items in DRO is owned by the author, with all rights reserved.

Every reasonable effort has been made to ensure that permission has been obtained for items included in DRO. If you believe that your rights have been infringed by this repository, please contact drosupport@deakin.edu.au.

Versions
Version Filter Type
Citation counts: TR Web of Science Citation Count  Cited 0 times in TR Web of Science
Scopus Citation Count Cited 0 times in Scopus
Google Scholar Search Google Scholar
Access Statistics: 56 Abstract Views, 1 File Downloads  -  Detailed Statistics
Created: Fri, 04 Oct 2019, 13:41:49 EST

Every reasonable effort has been made to ensure that permission has been obtained for items included in DRO. If you believe that your rights have been infringed by this repository, please contact drosupport@deakin.edu.au.