GIS Based Analysis and Accuracy Estimation of Sea Surface Temperature from MODIS Thermal Images

Capturing the thermal emission from the Earth, satellite sensors permit to derive sea surface temperature (SST). The infrared radiance intercepted by the sensor also depends on other variables, such as the surface emissivity, viewing geometry and the atmospheric contribution. All variables that affect atmospheric absorption and emission must be carefully considered. The brightness temperature of the surface measured by satellite sensor needs transformation to obtain SST. To remove the noise generated by the atmosphere, an equation can be used to derive a set of SST algorithm coefficients that can be applied to the brightness temperature image. Temperature data from specific points are necessary for the scope. This article aims to demonstrate that, in absence of an adequate temperature dataset, SST can be derived by MODIS thermal images (pixel dimensions: 1 km x 1 km) using Copernicus data even if it has a lower spatial resolution (0.0625° x 0.0625°). The study area includes an extended part of the Mediterranean Sea which contains the Strait of Sicily, the Adriatic, Ionian, and Tyrrhenian Seas. The Pathfinder algorithm developed by NOAA based on an equation with pre-established coefficients is also applied for comparison. Temperatures recorded by buoys are used to test the result accuracy. The experiments testify the good performance of the adopted approach.