Detection of land use/land cover changes for Penang Island, Malaysia

Abstract

One of the most significant challenges facing the Earth over the next century, most likely are land use/land cover (LULC) changes. This anthropogenic activity was caused by human, affect many parts of the Earth's system (e.g., carbon cycle, hydrology, and climate), global biodiversity, and the fundamental sustainability of lands. The monitoring of LULC changes along Penang Island, Malaysia is very important for the management, planner, and nongovernmental organizations since LULC data provide essential information for environmental management and planning. In addition, monitoring its LULC changes and to analysis the consequences of these changes provide valuable information for policy-makers, in order to support sustainable development. There has been growing interest in the use of remote sensing systems for a regular monitoring of the Earth's system. Obviously, there has been a great increasingly in significant, in order to understand the patterns of LULC changes over the last decade. Increased of the awareness of the issues resulted in the large number of focused studies directed to understand the nature of LULC changes. Improper planning in LULC changes can bring significant impacts on biophysical variables of the Earth's surface, such as land surface temperature (LST) and normalized difference vegetation index (NDVI). The aim of this study was to monitor the LULC changes in Penang Island, Malaysia. The standard supervised classification techniques (Neural Network, and Maximum Likelihood) classifiers were utilized, in order to extract thematic information from the acquired scene by using PCI Geomatica 10.3 image processing software. Multi-temporal Landsat images for the period of 1991–2009 were utilized, in order to examine LULC changes in Penang Island. The relative performances of the techniques were evaluated. The accuracy of each classification map was assessed using the reference data set consisted of a large number of samples collected per category. The study revealed that Neural Network classifier produced superior results and achieved a high degree of accuracy.