Change Detection Analysis of Urban Forest in Klang Valley using Multiemporal
Remote sensing Data: some Preliminary Result
Norhan Mat Yusoff and Mazian Hahin
Faculty of engineering and GeoInfromation science
Universiti Teknologi Malaysia
Locked Bag 791, 80990 Johar Bahru
Tel: 07-5502969, Fax: 07-5566163,
E-mail:mazlan@fksg.utm.my
Abstract
This paper reports some preliminary results of change analysis over kang Valley urban forests. In this study , a set of multitemporal Landsat Thematic Mapper data were transformed using Principal Component Analysis ( PCA) which the least correlated components produced were used in analyzing changes with in the forested area. Early results of this study indicated theta changes detected and analyzed periodically using remote sensing technique is useful in monitoring urban forest. Efforts rate now being made to translate the degree of changes detected to that of its nature on the ground.
Introduction
Development can be analysed by comparing the extent of growth periodically. Changes in land use activities and types influenced by the various types of land development. Economic and physical planning is inseparable as they are both major components in developing a nation. Klang Valley covers an area of 3,200 sq. Km. The rapid growth in population, industrial activities and developments are some of the major changes reflecting the continues 9% economic growth in the last 3 years. Some of the development are well monitored and detected by the authority. Undeniably, there are exploitations of urban forest for other uses in which these activities can be detected by collecting data periodically and analyzing any drastic changes. This paper reports changes that took place in the urban forest of Klang Valley, detected and analysed using remote sensing technique. Multitemporal remote sensing data namely Landsat Thematic acquired in 1991 and 1993 were used in this study, respectively. This paper reports the early study of change detection by using Principal Components Analysis ( PCA) to a single multitemporal dataset to produce higher components which are least correlated and related to change.
Material & Method
Study area
The region of study is the Klang Vallay and its surrounding area ( Fig. 1). The land cover class of interest is the urban. Land Thematic Mapper data of the year 1991 and 1993 were selected because of the improved radiometric , spectral spatial and temporal resolution by TM economic, land use and resource development. To understand and to map important landforms and their changes, we must have the capabilities to detect and map surface variability temporally, spatially and spectrally ( Chavez et. Al., 1996). Digital images were useful for monitoring and change detection because they allow area to be studied form a regional perspective, and the repeat coverage offered by satellites.
Figure 1: Study Area - Klang Vally is shown in the shaded area
