-
- C11B-03: Mapping ice stream sensitivity in the Amundsen Sector to changes in ice velocity observations (invited) (highlighted)
-
NOLA CC
Author(s):Generic 'disconnected' Message
Beatriz Recinos Rivas, University of Edinburgh (First Author, Presenting Author)
Daniel Goldberg, School of GeoSciences, University of Edinburgh
Noel Gourmelen, University of Edinburgh
James Maddison, University of Edinburgh
Ice-sheet models are key tools for predicting how much ice Antarctica will lose and how this will affect sea level rise. These models rely on satellite-derived ice speed data to tune unknown model parameters. Here, we used a fast method called Automatic Differentiation (AD) to map where -- and how much -- uncertainty/errors in satellite ice velocity observations affect model predictions. AD helps to quickly and accurately identify which observations have the biggest impact on our forecasts. Our results show that predictions are especially sensitive to errors in observations taken where the ice starts to float (grounding zone) and in areas influenced by tides. Additionally, we find that observations from farther upstream of the grounding zone can affect 40-year projections. This kind of analysis can guide future satellite data collection by showing which observations are the most important to improve. Our findings can help ice-sheet modellers know where to add other types of observations (radar and seismic observations or ice damage maps), to better estimate unknown parameters in areas where satellite velocity data is not enough to initialise the model and simulate ice stream dynamics.
Scientific DisciplineNeighborhoodType
Enter Note
Go to previous page in this tab
Session
