The Arctic Holocene Transitions Proxy Climate Database — Principal Millennial-Scale Patterns

Abstract:

The Arctic Holocene Transitions (AHT) Project is a community-based, PAGES-endorsed effort to investigate centennial-scale variability in the Arctic climate system during the Holocene, and to understand the feedbacks that lead to pronounced changes. The AHT project recently released a major database of Arctic Holocene proxy climate records (Clim. Past–Disc. 10:1). The systematic review of marine and terrestrial proxy climate time series is based on quantitative screening criteria with new approaches for assessing the geochronological accuracy of age models and for characterizing the climate variables represented by the proxies. Records from only 39% of the sites could be found in the primary paleoclimate data repositories, underscoring the importance of such community-based efforts to assembling a comprehensive product. The database authors, including representatives from six Arctic regions, considered published records from nearly 500 sites. Of these, time series from 170 sites met the criteria for inclusion in the database. Namely, the records are located north of 58°N, extend back at least to 6 cal ka (84% extend back > 8 ka), are resolved at sub-millennial scale (at least one value every 400 ± 200 yr) and have age models constrained by at least one age every 3000 yr. The database contains proxy records from lake sediment (60%), marine sediment (32%), glacier ice (5%), and other sources. Most (60%) reflect temperature (mainly summer warmth) and are primarily based on pollen, chironomid or diatom assemblages. Many (15%) reflect some aspect of hydroclimate as inferred from stable isotopes, pollen assemblages, and other indicators. Principal component (PC) analyses indicates that the predominant pattern of change in temperature-sensitive time series is a ramp between 5 and 3 ka that separates millennial-long intervals of less-pronounced change. This shift corresponds to cooling at most sites, but a substantial fraction of sites warm across this transition. Between 6 and 2 ka, 23% of the 124 sites warm (5% with p < 0.05), whereas 77% cool (47% with p < 0.05). The trend represented by the first PC is likely driven by orbital changes and shifting synoptic features; it does not capture other important features of the spatial-temporal pattern that are likely present in the database and that we are currently exploring.