BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Parameterising melt at the base of Antarctic ice shelves with a fe
 edforward neural network - Clara Burgard\, University of Grenoble Alpes
DTSTART:20230116T140000Z
DTEND:20230116T150000Z
UID:TALK194746@talks.cam.ac.uk
CONTACT:Dr. Shenjie Zhou
DESCRIPTION:The largest uncertainty when projecting the Antarctic contribu
 tion to sea-level rise comes from the ocean-induced melt at the base of An
 tarctic ice shelves. Current physics-based parameterisations used to link 
 the ocean temperature and salinity in front of ice shelves to the melt at 
 their base struggle to accurately simulate basal melt patterns. We explore
  the potential of a deep feedforward neural network as a basal melt parame
 terisation. To do so\, we train a neural network to emulate basal melt rat
 es simulated by highly-resolved circum-Antarctic ocean simulations. We exp
 lore the influence of different input variables and show that the neural n
 etwork struggles to generalise to ice-shelf geometries unseen during train
 ing\, while it generalises better on timesteps unseen during training. Thi
 s is work in progress and I am looking forward to discuss improvements and
  limitations of this approach.
LOCATION:BAS Seminar Room 1\;https://ukri.zoom.us/j/95564564065
END:VEVENT
END:VCALENDAR