LAND USE CHANGE ANALYSIS WITH MARKOV CHAIN [Case Study : Sukabumi Regency]

• Main Author: LIBRA FIRMAN SAHAT (NIM: 10114100), ALFONSUS
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/25352
• Institution: Institut Teknologi Bandung

Land is a place where humans carry out daily activities. The function of the land is as a place of residence, farm, industrial area, and so on. Helen Briassoulis (2012) said there are four models to model the potential for land change, one of them is a stochastic model. This research will use a stochastic model, more precisely the Markov Chain. Based on the parameter index, the Markov chain has a discrete and continuous time parameter index. This research will use the location parameter index to estimate the transition probability matrix. This is because according to geodesy theory, land change is not only caused by time but also the surrounding land. In addition, as a research area, Sukabumi Regency will be chosen which is the largest area in West Java. In estimating the transition probability matrix using the location parameter index, it will be viewed in one direction either vertically or horizontally from one location to another from the data of 2005 and 2010. After obtaining a one-step transition opportunity matrix, the transition probability matrix for n-steps will also be found as well as the classification of states by looking at the state is the land class. The results obtained from the one-step transition opportunity matrix and n-steps for 2005 and 2010 did not differ significantly after the χ^2 test. This result proves that for five years from 2005 to 2010 Sukabumi District land was not too different.