An Integrated Software Environment for Object-Based Cellular Automata: An Application to the Study of Land-Use Changes
Date
2019-09-12
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Cellular automata (CA) is a well-established modelling approach used to study patterns and dynamics of land-use/land-cover (LULC) systems and to predict their evolution. Increased computer performance, along with the need to improve how geographic space is represented have resulted in the recent development of object-based CA models. Their main advantage over conventional cell-based models is that they allow for the representation of meaningful, real-world entities. However, their use remains limited due to issues with data model inconsistencies between calibration and simulation, simple neighborhood configurations and driving factors, overlooking spatial and temporal scaling effects on simulated results, increased computation time required to handle vector geometrical transformations and topology, and lack of an integrated framework that encompasses the functionalities required for calibration and simulation. The objective of this research is to describe the architecture and functionality of a novel, object-based CA model that were tested in two study areas in the Elbow River watershed in southern Alberta at 5 m and 60 m resolution. A change detection analysis is first performed on a series of historical LULC maps in vector format to identify the trends and speed of change in LULC and the driving factors responsible for these changes. This information is stored in a spatial database accessible from the software environment. Calibration is conducted with several neighborhood configurations and drivers using the multi-class weight of evidence method to calculate the transition probabilities. Simulation is performed by allowing for the change of state and geometry of each object over time. Time-consuming vector-handling operations are optimized or parallelized to increase the speed of execution. The final model results indicate a positive agreement with an independent map used for comparison. The model reproduces realistic urbanization patterns along the main roads and the river. Also, it is apparent that there is a substantial improvement in computation time. This model represents a powerful exploration and application tool that will enable a large community of users to exploit the potential of CA modelling for understanding the dynamics of LULC systems.
Description
Keywords
Cellular Automata, GIS, Modeling, Land-use change
Citation
Amini Tareh, M. (2019). An Integrated Software Environment for Object-Based Cellular Automata: An Application to the Study of Land-Use Changes (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.