Quentin FEBVRE Thesis Defense

Tuesday 12.05.2023
Time:
From 09:00 to 11:00

Address:

IMT Atlantique, Campus de Brest, Salle B01-008A

Quentin FEBVRE, from the MEE department and the Lab-STICC laboratory, will present his research on the subject of :

Deep learning for ocean satellite altimetry : specificities and practical implications

Thesis defense Notice

 

See more

Abstract : This thesis explores how advances in deep learning can aid the analysis of satellite measurements of sea surface height (SSH). Current altimeters provide irregularly sampled data, limiting observation of the finest processes. Pushing back this limit would improve our climate monitoring capabilities. Exciting opportunities have emerged with the SWOT mission. Learning approaches have demonstrated remarkable capabilities in many areas. This thesis addresses the specific considerations of applying deep learning to altimetry data in three parts. Firstly, through the calibration of the KaRIn sensor, we demonstrate how domain-specific knowledge can be integrated into deep learning frameworks. Secondly, we address the scarcity of ground truth data when learning methods to interpolate altimetry data. We illustrate how simulations of ocean models and observing systems can overcome this challenge by providing supervised training environments that generalize to real data.Finally, our third contribution addresses the challenges encountered in bridging the gap between the "ocean" and "deep learning" communities. We describe how we tackled these aspects during the development of the OceanBench project.

Organizer(s)

As part of IMT Atlantique's thesis co-accreditation within the SPIN doctoral school

Key words : Deep learning, Altimetry, SWOT

Published on 29.11.2023
 
1
 
2
 
3
 
4
 
5
 
6
 
7
 
8
 
9
 
10
 
11
 
12
 
13
 
14
 
15
 
16
 
17
 
18
 
19
 
20
 
21
 
22
 
23
 
24
 
25
 
26
 
27
 
28
 
29
 
30
 
31