CLUSTER ANALYSIS-MARKOV CHAIN ON LAND COVER CHANGE DATA CASE STUDY: WEST JAVA PROVINCE
Land is a major factor that is closely related to human activity. In fact, there are many cases that human activities related to land use are uncontrolled. Furthermore, this condition causes land cover change and affect global climate. By knowing the pattern of land cover changes that occur, the cau...
Saved in:
Main Author: | |
---|---|
Format: | Theses |
Language: | Indonesia |
Online Access: | https://digilib.itb.ac.id/gdl/view/29679 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Institut Teknologi Bandung |
Language: | Indonesia |
Summary: | Land is a major factor that is closely related to human activity. In fact, there are many cases that human activities related to land use are uncontrolled. Furthermore, this condition causes land cover change and affect global climate. By knowing the pattern of land cover changes that occur, the cause of the phenomenon of land cover change can be better to understand. Because West Java Province has a fairly diverse physical condition, the pattern of change needs to be analyzed based on the division of its region. So, Cluster Analysis is applied. <br />
<br />
In general, cluster analysis is a method that is intended to reduce the dimensions of data on the object / individual. The principle of this method is to grouping objects based on their similarity. If between two objects have a large measure distance, so they are not similar and are in different clusters. In this case, cluster analysis is applied to clustering the region of West Java Province based on the physical condition of the land and its social economy. The measure distance that used is Euclidean considering the normality assumption of data are fulfilled. Because the land that is always in groups, the grouping technique used is a hierarchical method with a divisive approach. <br />
<br />
This divisive hierarchy method is applied to classify 4,242,837 objects based on the 13 variables. For the sample data of 40 objects that are taken systematically randomly obtained two large clusters. The cluster boundaries divide the Eastern (Cirebon, Majalengka, Ciamis, Tasik, and Pangandaran) and West of West Java province. In topographic side, the north and south are quite different. However, for the Northern and Southern regions bordering Central Java (eastern region) belong to the same cluster. It shows that the intensity of population density and the function of land applied by the government also play a role in the division of this region. It can be seen that the eastern part of West Java is classified as a place of tourism and the western part tends to be an industrial area. Furthermore, modeling of land cover changes with the Markov Chain method is performed by determining the transition probability matrix of each region. The pattern of change of cover for the eastern region is less likely to change than the western region. Changes in land <br />
<br />
classes within five years that occurred in the western region, among others, is forests into fields and fields into rice fields. |
---|