Thesis (M.Sc.) - Asian Institute of Technology, 2002

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science, School of Advanced Technologies

It is important to study the driving forces of land use change to understand the change process. Spatially explicit simulation models help to test hypotheses about landscape evolution under several scenarios. This research presents a dynamic simulation model of land use change of Nong Chok area, Central Thailand. Simulation of land use change has been performed integrating remote sensing, Geographical Information Systems (GIS) and dynamic simulation toolkit. The model has been developed based on selected biophysical and human driving forces. It is a cellular automata model that presents vicinity based transitional functions. The study was conceived for the simulation of land use change dynamics, in particular from paddy fields to fishponds. The model was run for 19 years from 1981-2000. Data describing present and historic land use pattern were derived from aerial photographs. Transition functions were developed following entropy calculation by using ID3 algorithm of the land use change datasets. The model uses as its input a land use map (1981), spatial and human variables: distance to canal, age, ownership, religion, education, and family size of the farmers before the simulation starts. The result of the simulation showed considerable performance of the model to diffuse fishponds except few mismatches. To validate this spatial simulation model of land use change dynamics, the simulated maps were compared with the reference land use map (2000) using a set of landscape indices: number of fishpond cells, patch density, mean patch size, edge density, fractal dimension, and mean nearest neighborhood. Further investigation by integrating other variables might allow the model to simulate land use change with greater accuracy.