Estimation of Land Surface Temperature over Reunion Island using the thermal infrared channels of Landsat-8

Abstract

Land Surface Temperature (LST) is one of key parameters in physics of land surface processes from local through global scale. LST is required for a wide variety of applications such as climatology, hydrology, agriculture, biochemistry, and change detection studies. The aim of this study is to estimate Land Surface Temperature using new Landsat-8 data. There are many available thermal infrared sensors but we choosed Landsat-8 data as it has a good spatial resolution. We use Split Window Algorithm which involves two parameters: emissivity and brightness temperature. This method corrects for atmospheric effects based on the differential absorption in adjacent infrared bands. For measurements in the atmospheric windows, the top of the atmosphere (TOA) radiance is then dominated by surface leaving-radiance. In this paper, the study area is Reunion Island. Elevations range from see level to 0 to 3070.50 m (Piton des Neiges). It has an area of about 200 km2. This island has southern hemisphere tropical climate. The image used is a multispectral image acquired by Landsat-8 OLI (Operational Land Imager-) and TIRS (Thermal Infrared Sensor) sensors on July 1st, 2013. We were able to establish LST maps of the study area. The results suggest that the methodology is feasible to estimate surface emissivity and surface temperature with reasonable accuracy over heterogeneous urban areas. Values observed on this map varied from 275 to 305°K (2°C to 32°C). We observe higher value of LST in the coastal zones and lower values are observed in mountainous area in the center of island.