Global warming due to loss of large ice masses and Arctic summer sea ice

The study addresses how the disintegration of major cryosphere components—the Arctic summer sea ice, mountain glaciers, the Greenland Ice Sheet (GIS), and the West Antarctic Ice Sheet (WAIS)—affects global mean temperature through climate feedbacks. Observations show rapid declines in Arctic summer sea ice (>10% per decade since the late 1970s), widespread retreat and mass loss of mountain glaciers, and accelerating mass loss from GIS and WAIS. Model studies suggest thresholds for substantial GIS loss between ~0.8 and 3.2 °C above pre-industrial, and parts of WAIS may already be unstable. Given potential near-term summer ice-free Arctic conditions and long-term ice-sheet changes—some with hysteresis and irreversibility—the paper quantifies the additional GMT increase due to loss of these elements and decomposes the contributions from albedo, lapse rate, water vapour, and clouds, focusing on radiative effects.

Prior work documents rapid Arctic sea-ice decline and potential near-future summer ice-free conditions under higher-emission scenarios, with some models indicating this even at moderate warming. Mountain glaciers have lost ~21% of their volume (1901–2009) with substantial committed future loss due to past emissions and disequilibrium. GIS and WAIS mass loss has accelerated, with expectations of further increases with warming; thresholds for GIS disintegration have been identified and WAIS instability has been observed in the Amundsen sector, potentially committing >1 m sea-level rise. Paleoclimate studies (e.g., Mid-Pliocene Warm Period) show strong polar amplification when large ice sheets were reduced. Previous estimates have explored radiative perturbations from Arctic sea-ice albedo changes (~0.3 W/m² for one month removal) and highlighted roles of temperature, albedo, cloud, lapse rate, and water-vapour feedbacks, with clouds likely contributing a positive feedback but with uncertainties. However, a comprehensive quantification of the additional GMT impact from concurrent large-scale cryosphere losses and feedback decomposition had been lacking.

The authors use the EMIC CLIMBER-2 with atmosphere, ocean, sea ice, vegetation, and land-ice components at coarse spatial resolution. They perform large ensembles of equilibrium simulations at fixed atmospheric CO2 concentrations from 280 to 700 ppm. A reference level of 400 ppm corresponds to ~1.5 °C above pre-industrial in CLIMBER-2. Experimental design: compare control runs (cryosphere elements intact) to perturbed runs where specific cryosphere elements are removed in an idealized manner to isolate their radiative feedback impacts. For GIS and WAIS, both ice cover and topography are removed; albedo is initially set to bare land or ocean and then allowed to evolve. For Arctic summer sea ice, the albedo of ice-covered areas is replaced by open-ocean albedo during JJA while retaining the thermodynamic role of sea ice to preserve energy and mass conservation; sea-ice area remains dynamically computed. Mountain glacier removal affects albedo but not topography. Isostatic rebound is neglected. Each simulation is integrated 10,000 years to equilibrium; diagnostics are averaged over the last 4000 years. Ensemble calibration: 39-member ensemble constructed by perturbing parameters influencing fast feedbacks (water vapour, lapse rate, clouds, albedo) to emulate GCM ranges (Soden & Held 2006). Constraints include combined water vapour+lapse rate feedback 0.8–1.2 W/m²/K, cloud feedback 0.3–1.1 W/m²/K, albedo feedback 0.2–0.45 W/m²/K, northern summer sea-ice minimum extent bounds, equilibrium climate sensitivity 2.0–3.75 °C, and small deviation at 280 ppm baseline. Feedback decomposition: top-of-atmosphere radiative perturbations are decomposed into albedo, cloud, and combined lapse rate + water vapour using the partial radiative perturbation method. The study focuses on purely radiative effects, excluding freshwater-induced dynamical feedbacks. Uncertainty is represented by ensemble medians and interquartile ranges of differences between perturbed and control equilibria.

- Total additional GMT from loss of Arctic summer sea ice (JJA), mountain glaciers, GIS, and WAIS at 400 ppm (~1.5 °C baseline) is 0.43 °C (0.39–0.46 °C), i.e., 29% (26–31%) extra warming relative to the 1.5 °C baseline. - Contributions by component (median, IQR): Arctic summer sea ice 0.19 °C (0.16–0.21), GIS 0.13 °C (0.12–0.14), mountain glaciers 0.08 °C (0.07–0.09), WAIS 0.05 °C (0.04–0.06). Effects add approximately linearly; the sum of individual removals matches removing all simultaneously. - Regional amplification: local warming up to ~5 °C above background near Greenland and West Antarctica; Arctic regional additional warming >1.5 °C annually; low-latitude warming ~0.2 °C. - Dependence on CO2: Additional warming from cryosphere loss is largely independent of background CO2 between 280–700 ppm except for Arctic summer sea ice, whose additional warming decreases as background CO2 (and thus baseline sea-ice loss) increases; observed late 20th-century minimum sea-ice area (~5.5–6.5 million km²) corresponds to ~0.15 °C additional warming in the model. - Feedback decomposition of radiative perturbations: ~55% from albedo changes; ~30% from combined lapse rate + water vapour; ~15% from clouds. For Arctic sea-ice summer removal, the albedo+cloud radiative perturbation is ~0.49 W/m², higher than estimates for a single month removal due to a longer low-ice period in these experiments. - Zonal/latitudinal structure: albedo and lapse rate+water vapour contributions are largest at high latitudes, reinforcing polar amplification; cloud feedback contributes less to polar amplification in these simulations.

The results demonstrate a significant additional commitment to global warming from the loss of major cryosphere elements beyond the direct greenhouse-gas-driven warming, with strong polar amplification and measurable warming globally including the tropics. On shorter timescales, a near-term loss of Arctic summer sea ice alone could add ~0.19 °C to GMT at a 1.5 °C baseline, while on longer centennial to millennial timescales, cumulative losses of GIS, WAIS, mountain glaciers, and Arctic summer sea ice add ~0.43 °C. The decomposition highlights that while prescribed albedo reductions dominate, non-albedo fast feedbacks (lapse rate, water vapour, clouds) account for over 40% of the additional radiative perturbation and thus are important contributors. The approximate linearity of contributions suggests that assessments may sum component-wise effects for first-order estimates. The findings align in magnitude and amplification patterns with paleoclimate warm periods (e.g., Pliocene). Importantly, some ice-sheet feedbacks may operate prior to complete disintegration, and threshold crossings for GIS/WAIS may be irreversible due to hysteresis. These results emphasize that preserving cryosphere elements can substantially limit additional long-term warming beyond nominal temperature targets.

The study quantifies the additional GMT increase from disintegration of key cryosphere components, finding ~0.43 °C extra equilibrium warming at 400 ppm (~1.5 °C baseline), with ~0.19 °C attributable to Arctic summer sea-ice loss alone. Albedo changes dominate the radiative perturbation, but lapse rate, water vapour, and cloud feedbacks together contribute over 40%. Contributions add approximately linearly across components. These findings underscore that protecting cryosphere stability is crucial to limit long-term warming and that climate targets should account for feedback-induced warming commitments. Future research should: integrate freshwater fluxes and associated ocean circulation responses; couple interactive ice-sheet dynamics and isostatic rebound; conduct transient simulations across timescales; refine cloud and high-latitude feedback representations; and evaluate results with higher-resolution GCMs and emergent constraints.

- Idealized experiments prescribe immediate removal/darkening of cryosphere elements (e.g., summer sea-ice albedo reduction) rather than simulating interactive, time-evolving ice dynamics. - Focus on purely radiative effects; freshwater fluxes from ice-sheet melt and associated ocean circulation feedbacks are excluded, potentially altering transient and regional responses. - Equilibrium, long-term (10,000-year) simulations may differ from transient, policy-relevant timescales; the additional warming quantified is a long-term commitment. - CLIMBER-2 is an EMIC with coarse spatial resolution and simplified cloud parameterizations; ensemble ECS range (2.0–3.75 °C) is narrower than full CMIP ranges, potentially narrowing response uncertainty. - No isostatic rebound; mountain glacier and Arctic sea-ice experiments neglect topographic effects; lapse rate and water vapour feedbacks are combined and not separated. - Results depend on calibration to specified fast-feedback ranges and on the representation of planetary albedo and clouds in CLIMBER-2.