Abstract: The lithosphere of Antarctica reacts to both the stress variations due to the dynamics of the mantle and the variation of the glacial load due to changes in climate. These two factors act on the dynamics of the glacial mass, on the evolution of the continent's topography and on the plate seismicity. This means that the knowledge of spatial variations in lithospheric thickness is needed to comprehend of the processes that involve interactions between climate and the geosphere. The main aim of this work is to produce a tomographic model of the continent. The classic data source for tomographic studies is a set of seismic signals produced by earthquakes. In this work I use a different technique: information is extracted by correlating the ambient seismic noise recorded in different locations. Instead of a the classic cross-correlation, this work uses the phase cross correlation technique, which appears to be more robust in our use case. Signal phases are used even to improve the correlogram stacks, by weighting the stack samples according to the correlograms' coherence. In Antarctica, the 30s Rayleigh wave group velocity map shows lower velocities in East Antarctica, and higher velocities in West Antarctica, coherently with expectations related to a thicker cratonic crust in the East, and thinner extending crust in the West. Deeper sensitivity of longer periods make the color pattern switch for the 80s map, sensing the upper mantle and consequently higher velocity at the cold roots of cratonic East Antarctica and lower velocity in the tectonically active West. I also contributed to the maintenance and development of the seismic observational infrastructure in Antarctica. This thesis recaps the activities I carried out during my participation to the 31st campaign of the Italian National Antarctic Research Program, to which I participated during the PhD course.