EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Quantifying uncertainty in future projections of ice loss from the Filchner-Ronne Ice Shelf System

Emily Hill1,2, Sebastian Rosier2, Hilmar Gudmundsson2, and Matthew Collins1
Emily Hill et al.
  • 1College of Engineering, Mathematics and Physical Sciences, University of Exeter, Exeter, United Kingdom
  • 2Department of Geography and Environmental Sciences, University of Northumbria, Newcastle-upon-Tyne, United Kingdom

Mass loss from the Antarctic Ice Sheet is the main source of uncertainty in projections of future sea-level rise, with important implications for coastal regions worldwide. Enhanced melt beneath ice shelves could destabilise large parts of the ice sheet, and further increase ice loss. Despite advances in our understanding of feedbacks in the ice sheet-ice shelf-ocean system, future projections of ice loss remain poorly constrained in many parts of Antarctica. In particular, there is ongoing debate surrounding the future of the Filchner-Ronne Ice Shelf (FRIS) region. The FRIS has remained relatively unchanged in recent decades, but an increase in air and ocean temperatures in the neighbouring Weddell Sea, could force rapid retreat in the near future. Indeed, previous modelling work has suggested the potential for widespread infiltration of warm water beneath the ice shelf in the second half of the twenty-first century, leading to a drastic increase in basal melting.

Here, we use the ice flow model Úa alongside the ocean box model PICO (Potsdam Ice-shelf Cavity mOdel) to understand the key physical processes and model variability in future projections of sea level rise from the FRIS region. We investigate uncertain model parameters associated with ice dynamics, surface melting and precipitation, ocean temperature forcing, and parameters relating to the strength of basal melt generated by PICO. We optimise the prior distributions of parameters in PICO using observations and a Bayesian approach, leading to improved posterior distributions for use in the following stages of uncertainty quantification. We then run our forward model through the 21st century for various RCP scenarios and extensive random sampling of uncertain parameters to train an emulator. From this, we present probabilistic projections of potential sea level rise from the FRIS region for different future climate change scenarios, together with a sensitivity analysis to identify the most important parameters that contribute to uncertainty in these projections.

How to cite: Hill, E., Rosier, S., Gudmundsson, H., and Collins, M.: Quantifying uncertainty in future projections of ice loss from the Filchner-Ronne Ice Shelf System, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-16045,, 2020


Display file

Comments on the display

AC: Author Comment | CC: Community Comment | Report abuse

displays version 1 – uploaded on 04 May 2020
  • CC1: Comment on EGU2020-16045, Nicolas Jourdain, 04 May 2020

    Hi Emily,
    I like your Bayesian approach to optimize PICO parameters. Do you find parameter values that are much different from the ones initially proposed by Resse et al ?

    • CC2: Reply to CC1, Emily Hill, 05 May 2020

      Hi Nicolas,

      Thanks, and good question! The point estimates for my pdfs give a lower value for the heat exchange coefficient and a similar/only slightly higher value for the overturning strength. In this case I’m optimising for four parameters (including a time delay and scaling coefficient that control the temperature in the outside box) rather than just the two PICO parameters. By including these additional parameters my point estimates also come out with a higher temperature than Reese et al. use. It would be interesting to repeat this for just the heat exchange coefficient and overturning strength and compare the results.