A Framework to Probe Uncertainties in Urban Cellular Automata Modelling Using a Novel Framework of Multilevel Density Approach: A Case Study for Wallonia Region, Belgium

Abstract

Urban expansion models are widely used to understand, analyze and predict any peculiar scenario based on input probabilities. Modelling and uncertainty are concomitant, and can occur due to reasons ranging from-discrepancies in input variables, unpredictable model parameters, spatio-temporal variability between observations, or malfunction in linking model variables under two different spatiotemporal scenarios. However, uncertainties often occur because of the interplay of model elements, structures, and the quality of data sources employed; as input parameters influence the behavior of cellular automaton (CA) models. Our study aims to address these uncertainties. While most studies consider neighborhood effects, timestep and spatial resolution, our study uniquely focuses on the susceptibility of multi density classes and varying cell size on uncertainty. Hence this chapter offers a theoretical elucidation of the concepts, sources, and strategies for managing uncertainty under various criteria as well as an algorithm for enumerating the model's accuracy for Wallonia, Belgium.

Modelling multi-density urban expansion using Cellular Automata for Brussels Metropolitan Development Area

“…Hence, in order to minimise the heavy computation time, we rasterised the data in 2×2m cell resolution. This data was aggregated into 100×100 m ( Mustafa et al, 2018b;Chakraborty et al, 2023;Tannier and Thomas, 2013) raster for creating density classes. All cell values < 25m 2 are considered non built-up because an average residential building of Belgium corresponds to values ≥ 25m 2 , represented by a cell size of 100m 2 in our aggregated data.…”

“…Three maps can define one calibration interval (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010) and one validation interval (2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020). The model considers a comprehensive set of 12 static causative factors encompassing accessibility, geophysical features, policies, socio-economic factors, and neighbourhood interactions, recognizing urbanization as a self-organizing system (Chakraborty et al, 2023).…”