Study on Land Surface Temperature Based on Landsat Image over Penang Island, Malaysia

Land surface temperature (LST) have been widely usedfor environmental studies and urban heat island (UHI),using satellite derived images. LST retrieval fromsatellite images has many advantages for research ascompared with in-situ measurement of air temperatures.For example, it will help to reduce the time consuming,more coverage of the study area with the high spatialresolution, etc. The objective of our study is to retrievethe LST over Penang Island and compare the trend andaverage value of LST within different types of (land useor land cover) LULC changes in single satellite image.LST was retrieved by ATCOR3_T using PCI Geomatica10.1 image processing software. The supervisedclassification was chosen to classified LULC changesover Penang Island. From the result obtained fromdifferent classification method, maximum likelihoodclassification showed the most accurate result comparedwith minimum distance-to-mean and parallelepipedmethods. The urban areas with high density resultedshown the highest LST between different LULC changes.These changes in LULC yield the significant differencein LST between urban and rural area. Therefore, LSTproved to be a good and effective tool to predict theimpact of LST to different LULC. Besides, therelationship between LST and NDVI also beeninvestigated for different LULC changes in this study.The strong negative and positive correlation wasobserved between LST and NDVI for different LULCchanges.