The Paris Climate Agreement and future sea-level rise from Antarctica

Abstract

The Paris Agreement aims to limit global mean warming in the 21 st century to less than 2 ºC above preindustrial levels, and to promote further efforts to limit the warming to 1.5 ºC. Here, we use an observationally calibrated ice sheet-shelf model including ductile and brittle processes that can initiate dynamic instabilities, to test Antarctica's response to future climate scenarios representing Paris Agreement aspirations versus more fossil-fuel intensive emissions scenarios. We find that global mean warming above 2 ºC substantially increases the risk of triggering rapid ice-sheet retreat, initiated by the thinning and loss of Antarctic ice shelves. A scenario consistent with current policies and allowing +3 ºC of warming by 2100 causes an abrupt jump in the pace of ice loss after ~2060, equivalent to ~0.5 cm sea level rise per year. Once initiated, rapid Antarctic ice loss continues for centuries, regardless of bedrock/sea level feedbacks or geoengineered carbon dioxide reduction (CDR). These results demonstrate the possibility that unstoppable, catastrophic sea level rise from Antarctica will be triggered if Paris Agreement temperature targets are exceeded. Greenland is currently losing ice at a faster pace than Antarctica 1,2 , but Antarctica contains almost eight (7.74) times more ice above floatation, equivalent to 58 m of global mean sea level (GMSL) 3 .The Antarctic Ice Sheet (AIS) is fundamentally different from the Greenland Ice Sheet, because most of its margin terminates directly in the surrounding ocean, with massive ice shelves (floating extensions of glacial ice) providing resistance (buttressing) to the seaward flow of the grounded ice upstream 4 . About a third of the AIS rests on bedrock hundreds to thousands of meters below sea level 3 and in places where subglacial bedrock slopes downward away from the ocean (reversesloped), the ice margin is susceptible to dynamical instabilities; the Marine Ice-Sheet Instability (MISI) 5,6 and possibly a Marine Ice-Cliff Instability (MICI) 7,8 that can drive rapid retreat. The West Antarctic Ice Sheet (WAIS), with the potential to cause ~5 m of sea level rise 3 , is particularly

Resolving GIA in response to modern and future ice loss at marine grounding lines in West Antarctica

“…Furthermore, this percentage decreased over time our simulations, and would continue to decrease multi-century and millennial simulations as the magnitude of viscous deformation and the scale of the ice loss continue to grow. While bedrock topography has smaller scale features (Morlinghem et al, 2020), our results suggest that the GIA signal is less localized and may be computed at lower resolution relative to the ice sheet dynamics and then interpolated and added to the initial topography on the higher resolution ice sheet model grid, as is done in, e.g., Gomez et al (2015) and DeConto et al (2021).…”

“…To investigate the influence of GIA model grid resolution, we first conduct idealized load sensitivity tests over a range of surface grid resolutions from 7.5 to 0.5 km, for the instantaneous removal of cylindrical loads from 0.5 to 16 km in radius (Section 3). We then widen our "aperture" to assess the model grid resolution required to accurately capture GIA due to modern ice-sheet cover changes from satellite observations (Shepherd et al, 2019) and future ice loss from ice-sheet model projections (Golledge et al, 2019;DeConto et al, 2021) in the rapidly evolving Amundsen Sea Embayment of West Antarctica (Section 4). We choose to locate our study region (light blue square in Fig.…”

“…For Antarctic evolution over the next century, we apply modelled ice thickness changes from two Antarctic-wide ice-sheet model projections: (1) ICE-GOL which predicts AIS evolution under RCP 8.5 and including meltwater feedbacks from 2000 to 2100 at 5-km resolution over 5-year intervals (Golledge et al, 2019), and (2) ICE-RD which predicts AIS evolution from 1950 to 2100 at 10-km resolution over the whole AIS with a nested 1-km resolution simulation over ASE at annual resolution (Extended Data Fig. 5 from DeConto et al, 2021). For ICE-RD, we ran simulations at yearly intervals from 1950-2100 but the interval from 1950-2000 is a period of ice model initialization and we therefore focus on the period between 2000 and 2100 in our results.…”

“…Alongside this, improvements in ice-sheet models (e.g. Nowicki et al, 2016;Seroussi et al, 2020;DeConto et al, 2021) and observational products (e.g. Studinger, 2014;Bamber et al, 2020;Smith et al, 2020;Morlinghem et al, 2020) allow similarly high-resolution km-to sub-km-ice thickness and bedrock topography datasets that serve as input to GIA models.…”

“…We repeat calculations with a range of Earth models, considering the contribution from elastic and viscous deformation separately. We focus on the response to observed modern ice loss over the last two decades (Shepherd et al, 2019) and projected future ice sheet retreat in the coming century (Golledge et al, 2019;DeConto et al, 2021) in the Amundsen Sea Embayment of West Antarctica. Our study is motivated by the following questions: What 3-D grid resolution is necessary to adequately capture the elastic and viscous uplift and associated gravitational and rotational effects in response ice loading changes?…”