Sustained North Atlantic warming drove anomalously intense MIS 11c interglacial

The fifth last glacial–interglacial transition (Termination V, ~428 kyr BP) and ensuing MIS 11c interglacial challenge classical orbital theory because a large-amplitude termination and one of the longest and warmest interglacials with high global sea level coincided with relatively weak boreal summer insolation (“MIS 11 paradox”). MIS 11c featured two insolation peaks (centred at 425.6 and 409.5 kyr BP). Prior evidence from the wider North Atlantic suggested sustained warm and/or humid conditions in the first half of MIS 11c, but chronologies relied on orbital tuning, far-field correlations, or sparse tephra layers, limiting robust evaluation of timing relative to insolation. This study aims to provide precise radiometric age constraints using 230Th-dated speleothem records from Bàsura cave (northern Italy), to assess the timing, sequence, and mechanisms of climate changes leading into and during MIS 11c, and to reconcile regional climate responses with insolation forcing and global ice-volume change.

Previous studies identified MIS 11c as an unusually long and warm interglacial with high sea level (6–13 m above present) despite weak insolation forcing, raising the “MIS 11 paradox.” Two insolation peaks occurred during MIS 11c, and recent work proposed caloric summer insolation thresholds governing termination timing, implying a later onset than some proxy evidence suggests. North Atlantic and Mediterranean records indicated early warmth/humidity in MIS 11c, but their chronologies were often orbitally tuned or correlated to Antarctic ice cores, precluding absolute timing tests. Model and proxy work pointed to strong AMOC and North Atlantic heat transport as drivers of regional warmth and Greenland ice loss during MIS 11c, but lacked precise synchronization with insolation and sea-level changes. Regional Mediterranean sensitivity to Atlantic climate via westerlies had been leveraged to study other terminations, highlighting the potential of high-resolution, well-dated speleothem archives to refine timing and mechanisms.

• Study site and sampling: Bàsura cave (44°08′N, 8°12′E, 200 m a.s.l.) in coastal Liguria, Italy. Flowstone core BA7-1 obtained in January 2014; targeted 317–1,120 mm interval.
• Geochemical proxies: 1,139 micromilled subsamples (10–50 µg) at 0.05–0.5 mm spacing for stable isotopes (δ18O, δ13C). δ18O and δ13C measured on Thermo-Finnigan MAT 253 (Fujian Normal University) and Micromass IsoPrime (National Taiwan Normal University); multiple international and in-house standards; reproducibility ±0.12‰ (δ18O, 1σ) and ±0.06‰ (δ13C, 1σ). Hendy tests performed on 14 layers to assess isotopic equilibrium.
• Trace elements: 558 subsamples (10–50 µg) at 0.1–0.5 mm spacing for Sr/Ca on Finnigan Element II ICP-SF-MS; 2σ reproducibility ±0.5% with matrix-matched standards.
• U–Th dating: 57 subsamples (0.4–4 g) dated by MC-ICP-MS (Thermo-Finnigan Neptune) with 229Th–233U–236U triple-spike; improved protocols for U isotopes; 54 ages used in age model; ages relative to 1950 CE with 2σ uncertainties. Age–depth modeling with StalAge and cross-checked with OxCal; StalAge chronology used.
• Proxy interpretation: δ18O and Sr/Ca interpreted primarily as precipitation amount and PCP extent, respectively, with δ18O also affected by temperature and source effects. Modern monitoring indicates drip response lag ~1 month; winter-half-year precipitation dominates recharge. Source water and meltwater inputs considered for δ18O shifts.
• Probabilistic analysis: Monte Carlo (5,000 simulations) to propagate chronological and proxy uncertainties, generate 68% and 95% confidence envelopes and medians for Bàsura, IODP Site U1313, and core MD03-2699; generated a probabilistic North Atlantic SST stack.
• Chronology transfer and synchronization: Two 230Th-based tie-points from maxima in 1-kyr filtered change rates (~425 and ~410 kyr BP) used to transfer Bàsura chronology to North Atlantic SST records (U1313, MD03-2699). Outside tie-points, original marine age models retained. Added ±3 kyr (2σ) to account for marine resolution and tuning uncertainty. Synchronization supported by alignment with Heinrich-like stadial/weak monsoon intervals.
• Onset detection: Applied RAMPFIT to SST records to identify transition from deglacial ramp to interglacial plateau; applied BREAKFIT to Bàsura δ13C. Independent threshold approach defined MIS 11c onset when 95% lower SST bound crossed local minimum Holocene SST (14.0 °C U1313; 16.7 °C MD03-2699). Sensitivity tests with varying search windows ensured robust change-point identification; errors from RMS propagation of chronological uncertainties.

• Two peak interglacial intervals in Bàsura: δ18O and Sr/Ca minima at 424–418 kyr BP and 412–402 kyr BP, separated by a ~6 kyr interval of higher values, indicating two warm/humid phases during MIS 11c.
• Radiometric onset at subtropical/mid-latitudes: BREAKFIT on Bàsura δ13C indicates onset of interglacial conditions at 423.1 ± 1.3 kyr BP (2σ), coincident (within uncertainties) with the first, weaker 65°N summer insolation maximum at 425.6 kyr BP.
• North Atlantic SST onsets (synchronized chronology): RAMPFIT onset at 425.0 ± 4.6 kyr BP (U1313) and 426.4 ± 4.6 kyr BP (MD03-2699). Holocene-threshold approach yields 425.5 ± 4.4 kyr BP (U1313) and 425.1 ± 4.4 kyr BP (MD03-2699). Results agree with original published age models (onsets ~424–426 kyr BP) and with Bàsura onset within uncertainties.
• Early regional interglacial versus global ice volume: At 423.1 ± 1.3 kyr BP, North Atlantic–Mediterranean climate was interglacial-like while global sea level remained at 41 ± 16 m (2σ) below present (Red Sea reconstruction). Sea level rose to present at 408.6 ± 2.5 kyr BP, then to +6–13 m, implying a lag of 14.5 ± 2.8 kyr between regional climate onset and attainment of global full-interglacial sea levels.
• Mechanism: A sustained (~15 kyr) period of strong AMOC and anomalously intense northward heat transport from subtropical latitudes during early MIS 11c preconditioned high latitudes, supporting extensive Greenland ice reduction and culminating in the MIS 11c maximum near the second, somewhat stronger insolation peak (~409.5 kyr BP).
• Deglacial sequence: Late MIS 12 aridity and hiatus at Bàsura; initial decay of southern ice-sheet margins under modest insolation increase (436.4–425.6 kyr BP) driving ~80 m sea-level rise from ~−120 m; establishment of interglacial-mode AMOC by early MIS 11c; prolonged warming enabling eventual global full-interglacial conditions after ~410 kyr BP.
• Regional coherence: Early MIS 11c warmth/humidity extended from subtropical/mid-latitudes to southern Greenland and Iceland; Mediterranean hydroclimate closely linked to mid-latitude North Atlantic SST via westerlies and storminess.

The study resolves the MIS 11 paradox regionally by showing that interglacial-like climate in the North Atlantic–Mediterranean began during the first, weaker insolation maximum, yet global ice volume remained large. Precise U–Th dating anchors regional proxies and synchronized North Atlantic SST records to an absolute timescale, demonstrating a ~15 kyr lead of regional hydroclimate/SST over global sea-level rise to typical interglacial levels. This supports a mechanism in which an early-AMOC interglacial mode sustained prolonged poleward heat transport, preconditioning polar regions and facilitating extensive Greenland ice loss before the second insolation peak drove the interglacial culmination. The findings underscore the role of duration and persistence of moderate summer warmth, integrated summer energy forcing, and ocean circulation in controlling the magnitude and timing of ice-sheet retreat, helping to reconcile weak instantaneous insolation forcing with an intense, long interglacial. The coherence with Asian monsoon weak phases and Heinrich-like stadials further integrates North Atlantic dynamics with global teleconnections.

High-precision 230Th-dated speleothem records from Bàsura cave, synchronized with mid-latitude North Atlantic SST reconstructions, show that MIS 11c interglacial conditions began regionally at 423.1 ± 1.3 kyr BP, in phase with the first weak insolation peak, and that full global interglacial conditions (sea level ≥ present) were achieved ~14.5 ± 2.8 kyr later. A sustained, strong AMOC-driven poleward heat flux prior to the MIS 11c optimum likely enabled anomalously intense warming and extensive Greenland ice loss, explaining the exceptional character of MIS 11c under weak insolation. These results provide an absolute, radiometrically constrained framework for the sequence of deglacial processes across Termination V and highlight the importance of the duration of moderate warming and ocean circulation in interglacial development. Future work should extend absolute dating across regions, refine synchronization without tuning, quantify AMOC strength and heat transport through multi-proxy and modeling, and test whether similar early low-latitude warming precedes other terminations.

• Interpretation of speleothem δ18O involves multiple controls (precipitation amount, source effects, temperature, PCP), with changing seasonal patterns and moisture sources through time; disentangling contributions remains uncertain.
• Many North Atlantic high-latitude proxy records discussed are not absolutely dated; synchronization relies on tie-points and tuning to Bàsura, introducing additional age uncertainty (marine resolution ~1.5 kyr; added ±3 kyr).
• Change-point detection depends on chosen algorithms and search windows, although sensitivity tests were performed.
• The mechanism invoking strong AMOC and poleward heat flux is supported by models and proxy coherence but direct AMOC reconstructions with absolute dating for this interval are limited.
• The study does not rule out roles of other feedbacks (e.g., vegetation–albedo) and does not claim uniqueness of early low-latitude warming to MIS 11c; generalizability to other terminations requires further testing.