The West Antarctic Ice Sheet (WAIS) plays a significant role in the rise of global sea levels due to its ongoing loss of mass. This phenomenon is primarily driven by its interaction with the Southern Ocean, particularly in the Amundsen Sea region. The continuous melting of ice shelves weakens their supportive role and leads to the acceleration of glacier flow towards the ocean. Potential irreversible retreat of WAIS glaciers poses a major threat as they hold enough ice to raise global sea levels by 5.3 meters.
Several studies have suggested that the Amundsen Sea has warmed due to atmospheric changes over the past century, which serves as a plausible explanation for the mass loss of the WAIS. However, detecting long-term warming trends in this area has been challenging as data has only been collected since 1994, and warming trends exhibit strong decadal variability.
The situation for the WAIS may worsen if the Amundsen Sea experiences further warming throughout the 21st century. Researchers have hypothesized that this warming could be influenced by future climate change and mitigated through the reduction of greenhouse gas emissions. Nevertheless, current projections of ice-shelf basal melting in the Amundsen Sea are not fully reliable due to limitations in underlying ocean models.
To adequately address the challenges posed by sea level rise, it is crucial to analyze various scenarios representing the best and worst cases of fossil fuel use, as well as intermediate cases. By understanding the dependence of ice loss on different climate scenarios, policymakers and global communities can better identify the unavoidable sea-level rise that they must adapt to, and the extent of ice loss that can still be controlled through reducing greenhouse gas emissions.
The time frame in which different scenarios diverge and the effects of chosen climate policies become apparent is also essential in sea-level rise predictions. Accurate projections are necessary in analyzing the impact of internal climate variability compared to anthropogenic forcing on future sea-level rise.
Table of Contents
In this study, an ocean model of the Amundsen Sea incorporating sea ice and ice-shelf cavities is utilized to create a series of future projections. The model, a configuration of the Massachusetts Institute of Technology general circulation model (MITgcm) ice-ocean model, is forced by the atmospheric output from the Community Earth System Model (CESM1) climate model. Previously published and validated for present-day and twentieth-century simulations, the model calculates ice-shelf basal melting and corresponding heat and freshwater fluxes. Notably, there is no coupling to an ice-sheet model, meaning the ice-shelf geometry remains constant.
The study introduces a bias-correction method for the simulated CESM1 ocean fields to allow for evolving ocean boundary conditions over time. This improvement takes into account remote changes in water masses that align with atmospheric forcing.
Five core scenarios are simulated:
- Historical Scenario (1920-2005): Based on observed external forcing, both anthropogenic and natural.
- Paris 1.5°C and Paris 2°C Scenarios (2006-2100): Aim to stabilize global mean temperature change at the specified thresholds relative to pre-industrial conditions, following the Paris Agreement’s goals.
- RCP 4.5 (2006-2080) and RCP 8.5 (2006-2100) Scenarios: Follow Representative Concentration Pathways for future anthropogenic forcing with medium and high fossil fuel use, respectively.
These four scenarios provide a robust range of possible climate mitigation outcomes, though it is important to note that containing global temperature rise to 1.5°C is becoming increasingly unlikely with current warming levels. Conversely, the RCP 8.5 scenario may be overly pessimistic due to limited fossil fuel reserves.
Two more scenarios with climatological ocean boundary conditions were added to assess the impact of remote water mass changes. Ensembles of 5-10 members with differing internal climate variability realizations in CESM1 were applied to each scenario. This approach is crucial for the Amundsen Sea, as Pacific modes of variability heavily influence the region and have likely contributed to historical trends.
The analysis focuses on key factors such as melting, ocean-driven melting, ice shelves, floating ice shelves, ice-shelf basal melting, ocean currents, winds, and ocean water to comprehensively examine the future of the Amundsen Sea region in varying climate situations.
Scenario Dependence of Warming
All future projections show a significant warming of the Amundsen Sea and increased melting of its ice shelves. The spatial distribution of trends can be observed for the Paris 2°C scenario, where mid-depth temperature (200–700 m mean) is the water that directly affects ice-shelf cavities. Trends in mid-depth temperature significantly correlate with trends in ice-shelf basal mass loss. However, actual basal mass loss will also depend on other factors not accounted for in the simulations, such as changes in ice-shelf geometry.
Future warming and melting trends are considerably stronger than historical ones, with ensemble mean future warming trends ranging from 0.8 to 1.4 °C per century compared with the historical mean of 0.25 °C per century. Even under the most ambitious mitigation scenario (Paris 1.5°C), the Amundsen Sea warms three times faster than in the twentieth century. Local atmospheric changes are the primary driver of Amundsen Sea warming, with remote ocean forcing playing a secondary role.
There exists a relatively large ensemble spread: future warming trends can vary by a factor of two depending on the phasing of internal climate variability. Regardless, the individual warming trends are significant for every ensemble member of every future scenario. In terms of Amundsen Sea warming, the Paris 1.5 °C, Paris 2 °C, and RCP 4.5 trends are all statistically indistinguishable, showing that different scenarios have similar outcomes for both warming and melting. Only RCP 8.5, the most extreme scenario, is distinct from the others. This suggests that climate mitigation has limited power to prevent ocean warming, which controls sea-level rise from the West Antarctic Ice Sheet (WAIS).
Although RCP 8.5 has a stronger warming trend than the other future scenarios, this difference is not apparent until mid-century. Timeseries of ocean warming in the core scenarios show that all future ensembles are largely overlapping and have similar ensemble means for most of the century. RCP 8.5 eventually diverges from the other ensembles around 2045. This indicates that while mitigating the worst-case climate change scenario has the potential to reduce Amundsen Sea warming, it might not make a significant difference for several decades. By that time, the impact on some glacier basins of the WAIS could be irreversible, even if ocean temperatures then returned to present-day values.
Future scenarios are expected to diverge more in the twenty-second century and beyond. Although Paris 1.5 °C is not distinct from the two mid-range scenarios when considering trends over the full period, its warming trajectory noticeably flattens out towards the end of the simulation. By that time, the underlying climate model requires net negative CO2 emissions to stay below 1.5 °C of global warming.
Mechanism of Warming
The processes responsible for ocean warming can be inferred by examining the vertical temperature structure of the water column. On the continental shelf, there is a seasonal surface layer, a year-round subsurface cold Winter Water layer (WW, about -1.5°C), and a warm Circumpolar Deep Water (CDW, approximately 1°C) layer beneath it. These layers are separated by a sharp temperature gradient known as the thermocline, located around 100-400 meters.
In future scenarios, the thermocline rises, leading to increased volume of warm CDW reaching higher depths in the water column. In moderate warming scenarios, the thermocline’s position exhibits some variability, resulting in fluctuations between warm and cold conditions in the ice-shelf cavities. However, in extreme warming scenarios, such as RCP 8.5, the thermocline rises to a level where its variability no longer affects the cavities, causing them to be consistently exposed to warm water.
Ocean warming on the continental shelf is primarily driven by an increased volume of CDW rather than an increase in CDW temperature itself. This rise in thermocline and influx of warm water results from intensified circulation over the continental shelf and slope, bringing more CDW onshore. As the Amundsen Undercurrent, a bottom current transporting CDW eastward along the shelf break, strengthens, onshore transport of CDW is also enhanced, particularly in the Pine Island Thwaites East Trough (PITE) and Dotson-Getz Trough.
Increased CDW flux leads to the warming of the continental shelf and adjacent ice-shelf cavities. Trends in continental shelf temperature strongly correlate with southward transport trends through the PITE Trough, indicating that the intensification of the Amundsen Undercurrent is the primary driver of Amundsen Sea warming and ice-shelf melting.
The atmospheric drivers behind these oceanographic changes appear to be connected to surface warming and increased precipitation, rather than winds as previously suggested. Nonetheless, a dominant driver explaining the variability in trends within each ensemble has yet to be identified.
Overall, the mechanism of warming in the Amundsen Sea region is dictated by the interaction of oceanographic and atmospheric processes, resulting in the rising of the thermocline, increased volume of warm water intruding onto the continental shelf, and amplified ice-shelf melting rates due to ongoing climate change and global warming.
Relevance to Sea-Level Rise
Increased melting at the base of ice shelves can lead to a loss of buttressing, which in turn can cause more ice mass to flow across the grounding line, ultimately contributing to sea-level rise. Though it is challenging to quantify the exact contribution to sea-level rise based on ocean simulations alone, the significance of this can be indirectly assessed by examining the spatial distribution of basal melting trends.
The concept of the buttressing flux response number (BFRN) quantifies the potential impact of ice-shelf basal melting on sea-level rise. Regions with higher BFRN values are more likely to contribute to sea-level rise as they undergo increased basal melting. Significant regions with high BFRN values include the grounding lines of most ice shelves and the shear margins of Pine Island and Thwaites ice shelves.
Assessing the future scenarios of sea-level rise, it becomes evident that increased melting affects all BFRN classes, especially the most crucial ones with a BFRN greater than 10%. This indicates that future melting projections do not disproportionately impact either low or high BFRN regions. However, the historical melting trends are concentrated in high BFRN classes.
The main findings from ocean warming, such as higher ensemble mean trends in the worst-case scenario (RCP 8.5) compared to lower-range scenarios, also apply to buttressing implications. Interestingly, the additional increases in melting in RCP 8.5 are predominantly among ice-shelf classes that have less potential to cause sea-level rise. This is due to the additional thermocline rise in RCP 8.5, primarily affecting shallower ice drafts with generally lower BFRN values. Consequently, deeper ice with higher BFRN values becomes submerged by Circumpolar Deep Water (CDW) in all future scenarios.
Therefore, even mitigation efforts for the worst-case scenario might not substantially reduce the future sea-level rise contribution from the West Antarctic Ice Sheet (WAIS). The ongoing melting of ice shelves, particularly in high BFRN regions, poses significant threats to coastal cities worldwide as sea-level rise continues to accelerate due to climate change.
The simulations indicate a concerning future for the Amundsen Sea, as significant ocean warming and ice-shelf melting are expected in all future climate scenarios. This includes even the most ambitious climate goals, suggesting that rapid ocean warming and sea-level rise may be unavoidable this century. The primary cause of this warming is the accelerated transport of warmer Circumpolar Deep Water (CDW) onto the continental shelf by the Amundsen Undercurrent.
Both mid-range emissions scenarios (RCP 4.5) and the ambitious Paris Agreement’s targets (limiting global warming to 1.5°C or 2°C) show similar warming trends for the Amundsen Sea throughout the 21st century. This suggests that internal climate variability is crucial in determining the fate of the West Antarctic Ice Sheet (WAIS). Mitigation efforts may only prevent the worst-case scenario (RCP 8.5), where the thermocline is so high that most ice shelves are constantly exposed to warm CDW.
This research presents the most comprehensive future projections of Amundsen Sea ice-shelf melting to date, simulating a wide range of scenarios with ensembles for statistically robust comparisons. Combining the maximum future warming trend in the ensembles with historical warming reveals that ocean conditions in the Amundsen Sea could be up to 2°C warmer than pre-industrial temperatures by 2100, which is significant for Antarctic water masses.
However, further research and model development are needed to increase confidence in these conclusions. Currently, this study uses a single ice-ocean model forced by a single climate model and does not account for feedbacks related to ice-shelf geometry or long-term atmospheric forcing on the ice sheet’s surface mass balance. These factors could introduce a stronger sensitivity to the climate forcing scenario.
Nonetheless, this study does not diminish the importance of mitigation efforts in limiting the impacts of climate change. Mass loss from the WAIS is just one component of sea-level rise, and other regions of Antarctica may not lose substantial mass if current emissions targets are met. Climate change has numerous consequences beyond sea-level rise, making both mitigation and adaptation essential strategies in the global response to these challenges.
The opportunity to preserve the WAIS in its current state may have already passed, and policymakers should prepare for several meters of sea-level rise in the coming centuries. Internal climate variability, which is unpredictable and uncontrollable, might be the decisive factor in determining the rate of ice loss during this time. To minimize the societal and economic costs of sea-level rise, efforts must be focused on a combination of mitigation, adaptation, and luck.
The study utilized the MITgcm6 model, which focuses on the Amundsen Sea region with a high-resolution of approximately 3-5 km. The model comprises ocean, sea-ice, and ice-shelf thermodynamics modules. The simulations applied CESM1 climate model experiments as their forcing. The MITgcm’s performance was validated against the observed continental shelf temperature and salinity, as well as the Amundsen Undercurrent. Several future scenarios were generated, following different forcing pathways like Paris 1.5°C, Paris 2°C, RCP 4.5, and RCP 8.5.
Open Boundary Bias-Correction Technique
Since global climate model CESM1 has relatively coarse resolution and is not optimized for the Antarctic continental shelf, biases in water mass properties are addressed through bias-correction methodology. This approach uses ‘normalized T/S space’ to index water masses based on their relative temperature and salinity. It preserves CESM1’s transient changes in water mass properties without adopting mean-state biases.
Bias-correction consists of these main steps:
- Transform raw CESM1 output fields to normalized T/S space
- Define temperature and salinity anomalies for every possible bin
- Correct biases in each bin with non-zero volume
- Convert bias-corrected values back to the physical space using World Ocean Atlas water mass distribution
- Implement special treatment for the mixed layer to ensure consistency with atmospheric forcing
This methodology effectively maintains trends in CESM1 water mass properties and structure while retaining the more accurate water mass structure of the World Ocean Atlas.
Statistical Conventions Implemented
To analyze the data, a subset of available ensemble members from CESM1 was used. The core Historical experiment included ten members, whereas five additional members were incorporated for Historical Fixed BCs. The future scenarios had ten members each, except for Paris 1.5°C which contained only five members due to data limitations. This approach ensured statistical robustness while managing computational costs and data availability.
Ice-Shelf Buttressing Model Simulations
A critical component of the study was simulating ice-shelf buttressing, which refers to the interaction between ice shelves and adjacent grounded ice. These simulations were carried out to analyze how melting ice shelves may affect sea level rise and overall stability of the West Antarctic Ice Sheet.