Exploring frontogenesis processes in new satellite sea surface temperature data sets

Abstract

This PhD thesis is about understanding some aspects of the dynamics of the ocean surface mixed layer by means of satellite Sea Surface Temperature (SST) measurements. The focus is on surface density fronts which are used as a measure of the dynamical activity at the meso- and submesoscale. A review of the current knowledge on the non-linear physical mechanisms that occur in the vicinity of fronts is presented in Chapter 2. Chapter 3 characterizes the ability of two algorithms for detecting fronts of various sizes and strengths that are embedded in a complex turbulent flow and sampled on noisy two-dimensional images. Chapter 4 explores and compares the performances of several new multi-sensor Level-4 SST products in resolving the small scale gradients. These Group for High Resolution Sea Surface Temperature (GHRSST) data sets offer an unprecedented spatial and temporal SST coverage. Their production however involves a variable and partially unknown level of smoothing which “hides” some of the small scale variability. In Chapter 5, a robust, flexible, automatic and optimized Matlab-based methodology for detecting fronts on SST images and calculating a frontal index is presented. A basic frontal index capturing the frontal length and strength is exploited to quantify the spatial scales present in the various Level-4 SST products. More advanced frontal indexes based on physical oceanography results by others are constructed in order to estimate vertical exchanges occurring at fronts from their signature on the SST. These new frontal indexes, which characterize fronts according to their dynamical significance, allow the quantification of the upwelling, subduction and restratification associated with frontal submesoscale processes. Finally, the spatial and temporal variability of ocean fronts is explored in order to determine their sensitivity to climatic signals.

More information

Identifiers

Catalogue record

Export record