Abstract:

Build - up areas and bare land data are crucial for studying urbanization and environmental monitoring. Recently, remote sensing has become an effective tool for mapping these objects. The conventional approach is to classify images, however the use of index images produces faster and more accurate results. Previous studies using index images showed that the results obtained depend on the climate region, the parammetters of sensor and the features of each city; In which the original indicators of NDBI, IBI, EBBI, NBI, UI, MNDISI are proposed for Landsat TM, ETM + images and mainly in other climate zones, the number of studies for tropical region such as Vietnam is still limited. Therefore, this study examines the feasibility of these indices to new generation satellite - Landsat 8 for mapping build-up areas and bare land in tropical urban of Hanoi. The results showed that the NDBI is the most relevant and accurate, followed by IBI, EBBI, NBI, UI and finally MNDISI with overall accuracy of 87.25%; 86.96%; 85.25%; 82.89%; 77.24%; 75.68%, respectively; This study also pointed out differences in the results compared to prior studies and provided data on bare land and build-up areas for urban heat island monitoring in Hanoi.