Integration of multi-agent simulation (MAS) techniques with remote sensing for coastal land use change modeling

• Main Author: Paringit, Maria Cecilia Rubio
• Format: text
• Published: Animo Repository 2002
• Subjects: Coasts—Remote sensing; Land use—Remote sensing; Civil and Environmental Engineering
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/13541
• Institution: De La Salle University

This paper describes the development as well as the essential features of a land cover use change simulation model in coastal areas based on the principle of multi-agent systems with specific emphasis to the application of remotely sensed data. A multi-agent system (MAS) is a set of agents or "actors" interacting in a common environment, modifying themselves and their environment in the process, and thus able to explain changes in land use especially in coastal zones where enormous environmental transformation takes place due to both anthropogenic and natural causes. In this paper, two stages of image analysis have been examined in order to establish the viability of MAS for change detection. First, a set of Landsat multi-temporal images was analyzed in terms of spectral responses from different land cover types. Actors responsible for changes detected were identified together with their associated transition rules in land cover or use change. The transition rules were organized to come up with a simulation model that produces the spatial distribution of land uses through time given a set of agent and knowledge of their apparent decision. The images eventually classified according to change were used to validate the results of the simulation model. This effort holds the potential to contribute in modeling the complex process of land use planning by providing a better understanding of actor-driven influences to modifications of the natural environment.