Skip to Main content Skip to Navigation

Apport de l'altimétrie radar spatiale à l'étude de la neige de la calotte polaire Antarctique

Abstract : The Antarctic ice-sheet is composed by a thick layer of ice surmounted by firn of up to 100 m thick. For now almost 40 years, the satellite observations provide a great number of information on its topography and some properties of the snow surface. However, the knowledge of its shallow and deeper snowpack properties is still very limited. The snow precipitations are transported and deposed by the wind to form layers. The successive accumulations of snow bury the snow layers, which are modified under the temperature and pressure conditions. Yet, the accumulation rate or the snow grain size are important climate indicators, not well known, whom signatures are contained in the near-surface observations.

Since 2002, the ENVISAT satellite provides altimetric data over 80 \% of the Antarctic ice-sheet simultaneously at 2 different frequencies (S band at 3.2 GHz and Ku band at 13.6 GHz). These two waves penetrate into the snowpack over several meters and are reflected differently by its different elements, depending on the frequency. The idea of this thesis is therefore to use the dual frequency information to retrieve properties of the snowpack.

We solve this problematic by an analysis and a modelling of the dual frequency altimetric signal over the ice-sheet, and then by its inversion. We focus first on different case studies to estimate the sensitivity of these signals to different properties of the snow: i/ the altimetric signal is sensitive to the snow surface roughness at different scales, ii/ the altimetric signal varies seasonally under compaction of the near-surface snow, iii/ the radar waves are reflected by deep layers of ice or snow.

A model of altimetric signal is developed to compare the two frequency datasets over the whole ice-sheet. The results of the model are used to explain the seasonal variations previously observed. These results provide the first observations of the seasonal compaction of the snowpack. Finally, the snowpack properties are estimated at a global scale over the Antarctic ice-sheet. The snow grains are found to be larger inside the continent than on the coast. The snow density shows seasonal variations of up to 0.15, with greater amplitudes near the coasts. Some regions (Dronning Maud Land for instance) present a very smooth snow surface aspect.

The in situ measurements of the snow surface roughness are almost inexistent on ice-sheets, so that we develop a new method to measure snow surface profiles based on a distance-sensor laser onboard a skidoo. This instrument is successfully tested on the Midre Lovenbreen glacier in Svalbard.
Document type :
Complete list of metadatas

Cited literature [120 references]  Display  Hide  Download
Contributor : Williams Exbrayat <>
Submitted on : Thursday, January 24, 2008 - 2:52:08 PM
Last modification on : Monday, February 22, 2021 - 9:46:02 AM
Long-term archiving on: : Thursday, April 15, 2010 - 4:08:34 PM


  • HAL Id : tel-00216105, version 1



Pascal Lacroix. Apport de l'altimétrie radar spatiale à l'étude de la neige de la calotte polaire Antarctique. Océan, Atmosphère. Université Paul Sabatier - Toulouse III, 2007. Français. ⟨tel-00216105⟩



Record views


Files downloads