Antarctic Sea Ice Variations and Their Linkages with Global Extreme Weather Events

Abstract. Antarctic sea ice is a critical indicator of global climate dynamics, yet its post-2015 accelerated retreat and links to extreme weather/key climate modes remain insufficiently characterized. To address these gaps, we analyse 2010–2024 multi-source data (NSIDC sea ice extent (SIE), CMA extreme weather records, NOAA ENSO indices) via spatiotemporal decomposition, non-linear modelling, and mechanistic dissection.​ Antarctic SIE showed a "stable-then-decline" trend: minimal variability 2010–2014 (peak: 20.16 × 10⁶ km², Sep 2014), followed by unprecedented post-2015 retreat to a Feb 2023 historic minimum (1.85 × 10⁶ km², ~20 % below 2010–2014 summer mean). The Amundsen–Bellingshausen Sea and Antarctic Peninsula were main retreat zones, with 2020–2024 declines (10–15 %/20–25 % winter/summer) exceeding East Antarctica (5–10 %) and Ross Sea (8–12 %).​ We identified tiered negative covariance between SIE and extreme weather, presumably co-modulated by the El Niño-Southern Oscillation (ENSO) and Southern Annular Mode (SAM): strong associations (R² ≥ 0.6, P < 0.001) for extreme heat (400 % frequency increase) and cold waves (700 % increase), potentially via albedo–circulation feedback; moderate associations (0.5 ≤ R² < 0.6, 0.001 ≤ P < 0.01) for floods/rainstorm/typhoons, likely from ACC heat transport changes and convective propagation; weak associations (R² < 0.4, P ≈ 0.05) for blizzards, possibly due to spatially constrained El Niño–SAM effects. La Niña-positive SAM may have amplified these linkages (e.g., 400 % extreme heat increase in 2021–2023 triple La Niña), while El Niño-negative SAM suppressed them. These findings advance polar-low latitude coupling understanding, aiding extreme weather prediction and IPCC AR6-aligned adaptation.​