Current tracking in the Mediterranean Sea using thermal satellite imagery

Abstract

Infrared images representing the thermal state of the ocean surface may be used to track ocean currents with feature-tracking algorithms that identify and follow temperature gradients and hence represent advective surface motion. One of these techniques, the Maximum Cross Correlation technique, is based on a comparison of individual subscenes of sequential images, to estimate where a feature has moved from one image to the next. It has distinct advantages compared to alternative feature tracking algorithms such as its simplicity and robustness. Previous research has shown it to achieve a precision of 0.08 to 0.20 m/s rms. This study focuses on surface currents in the central Mediterranean Sea by analysing sequential Advanced Very High Resolution Radiometer local area coverage 1.1 km resolution images from June 2003. Most attention is placed onto the presentation of the method, the results and the main ad/disadvantages of using a Maximum Cross Correlation approach. © 2008 Springer Science+Business Media B.V.