A FULL HOLOCENE SUMMER TEMPERATURE RECONSTRUCTION FROM PRECIPITATION ISOTOPES IN SYNGENETIC PERMAFROST IN CENTRAL YUKON TERRITORY (EASTERN BERINGIA)

The Holocene temperature history of eastern Beringia is mostly informed by biological proxies, including fossil pollen and midges, which offer conflicting accounts on the timing and style of early Holocene peak warmth. Precipitation isotope archives such as ice cores offer an inorganic perspective on past temperatures in High Arctic locales, but are absent in eastern Beringia. Here we present a novel 13 ka isotope record from pore ice in syngenetic permafrost from a soligenic peatland in central Yukon. Ice-rich peat cores were collected with a portable drilling system, and stable water isotope ratios of the pore ice (δD_ice and δ¹⁸O_ice) were measured at a mean sampling interval of ~240 years. AMS ¹⁴C dating of peat fibres constrains the pore ice chronology. The pore ice record integrates summer precipitation that reaches maximum thaw depth in summer, freezes in situ during fall freezeback and is archived permanently by aggrading permafrost paced by peat accumulation. Pore ice is isotopically consistent with local summer precipitation, which is closely linked to air temperatures (3.48‰·°C^-1) in the region. We use this relation to estimate past temperatures from δD_ice, after correcting for δD_seawater changes. Our reconstruction reveals a protracted warm phase from 10.3-6.8 ka BP, with optima that were ~0.3-0.4°C warmer than the Holocene average. This warm phase was followed by a ~5 ka cooling leading to mid-Common Era temperatures that were ~0.4°C below the Holocene mean. This cool phase was countered by rapid 20^th century warming. Top of permafrost δD_ice is ~ +1.5‰ compared to the early Holocene climate optima pore ice, suggesting modern temperatures are the warmest on record by a ~0.4°C margin. Our reconstruction is remarkably coherent with N. Hemisphere multi-proxy temperature compilations, and explained by orbital and CO₂ forcing. Locally, our reconstruction of the early Holocene is more consistent with midge-based estimates than pollen-based estimates, which may reflect important seasonal biases between the two biological proxies. Our reconstruction provides novel insights that help advance a long-standing debate on the Holocene temperature history of eastern Beringia, and demonstrates great potential to develop syngenetic pore ice records in other permafrost regions where ice core records are not available.