(Table S1) Age determination of sediment core RAPiD-21-03K

A decadal resolution time series of sea surface temperature (SST) spanning the last two millennia is reconstructed by combining a proxy record from a new sediment sequence with previously published data from core MD99-2275, north of Iceland. The alkenone based SST reconstruction is validated with historic observational data and compared to a new similar temporal resolution reconstruction obtained from sediment core RAPiD21-3K, in the subpolar North Atlantic. The two SST paleorecords show consistent multidecadal scale coolings throughout the interval and similar expressions during the contrasted climatic periods 'colloquially known' as the Medieval Climatic Anomaly (MCA) and Little Ice Age (LIA). In order to further understand the temporal and spatial SST variations and investigate the influence of natural forcings on the observed SST changes during the last millennium, we compare our time series to simulations using the Institut Pierre-Simon Laplace IPSLCM4-v2 climate model. This comparison highlights the potential importance of volcanism as a natural forcing driving coherent abrupt cooling events captured in the subpolar North Atlantic records.

(Table T1) Ages of white mica clasts in core samples of ODP Hole 210-1276A

Fifty-seven white mica clasts were separated from five samples taken from near the bases of turbidites ranging in age from early Albian to middle Eocene. Twenty two (39%) of the micas have ages between 260 and 340 Ma and five (9%) have older ages (~400-600 Ma). The former age range is characteristic of the North American Alleghenian orogeny and the Iberian Variscan orogeny. The latter range is characteristic of the North American Acadian orogeny and older basement rocks in the Grand Banks and Newfoundland areas. Both age ranges are present in the middle Eocene sample, but only the younger range occurs in the middle Albian sample. This difference could be a sampling artifact. If this is not the case, then the most likely explanation is that the Acadian-aged micas within the Meguma Zone underlying the Grand Banks were totally reset by Alleghenian reactivation of the zone, a feature which occurs extensively in Nova Scotia. The addition of Acadian-aged micas in the middle Eocene sample may reflect a change in sediment provenance as drainage systems unrelated to rift topography developed. With the exception of one clast dated at 186 Ma, the 12 other micas obtained from the upper Paleocene sample yielded ages between 55 and 74 Ma, with 7 falling within ±2 m.y. of the 57-Ma age of the sample indicated by the biostratigraphic age-depth plot for Site 1276. This, together with the volcaniclastic content of the sample, indicates an input from near-contemporaneous volcanism. The nearest known occurrences of near-contemporaneous late Paleocene volcanism that could have produced white micas are in Greenland and Portugal, some 2000 and 1500 km distant, respectively, from Site 1276 during the Paleocene. However, ages of volcanism in these areas indicate that they could probably not be sources of micas younger than 60 m.y., which suggests some as-yet unknown volcanic source in the North Atlantic area. Accumulation in the Grand Banks area of airborne-transported volcaniclastic material from eruptions of slightly different ages, followed by a single resedimentation event, could account for the spread of dates obtained from the sample. White micas from the lowermost Albian sample show a spread of ages between 37 and 284 Ma that is completely different from the age distribution pattern of the middle Albian and middle Eocene samples. The sample location is between, and at least 25 m above and below, two igneous sills dated at 98 and 105 Ma. The sills have narrow thermal aureoles and ages older than the youngest detrital micas in the sample. It is unlikely, therefore, that the spread of mica ages in the sample is due to partial resetting of ages caused by thermal effects associated with the intrusion of the sills. The resetting may have been associated with a longer lived thermal event.

Geological map of Potter Peninsula (King George Island, South Shetland Islands, Antarctic Peninsula)

We present here a new geological map of Potter Peninsula (King George Island, South Shetland Islands). Like on adjacent Barton Peninsula, the morphology on Potter Peninsula is predominantly characterized by a glacial landscape with abrasion platforms offshore, in parts steep cliffs along the coast, and a rather smooth, hilly countryside in the interior. Potter Peninsula forms part of the downthrown Warszawa Block. The volcanic sequence cropping out here belongs to the King George Island Supergroup, with an observed local minimum thickness of approx. 90 m (Kraus 2005). The most prominent morphological feature is Three Brothers Hill (196 m), a well known andesitic plug showing conspicuous columnar jointing. It marks the final stage of activity of a Paleogene volcano, whose eruption products (lava flows and pyroclastic rocks), together with hypabyssal intrusions related to the volcanism, make up most of the lithology observed on Potter Peninsula (Kraus 2005). The Three Brothers Hill volcanic complex is eroded down to its deepest levels. Thus, the stratigraphically deepest units from the initial phase of volcanic activity are cropping out in some parts (Kraus & del Valle, in Wienke et al. 2008). The lithology on Potter Peninsula comprises lava flows (~50%), pyroclastic rocks (ash-fallout, pyroclastic flow deposits, volcanic breccia and agglomerates, ~30%) and hypabyssal intrusions (dykes, sills and small subvolcanic intrusive bodies, ~20%). 40Ar/39Ar datings carried out on magmatic dykes from Potter Peninsula indicate a short, but intense intrusive event during the Lutetian (Kraus et al. 2007).

Feldspar size measurements of tephra layers from ODP Leg 165 sites

Crystal size measurements have been carried out on tephra fall layers of Miocene to recent age from Sites 998, 999, and 1000 in the western Caribbean Sea. Maximum crystal size is used as a proxy for the grain size characteristics of the layers and an index of atmospheric dispersal from source eruptions. Crystal sizes range from 50 to 650 µm with the majority falling between 200 and 300 µm. All three sites exhibit a coarsening in the grain size of tephra layers with increasing age to the early Miocene that broadly correlates with an increase in the frequency of layers. Analysis of the present lower and upper level atmospheric circulation in the western Caribbean suggests that the layers were derived from source eruptions to the west of the sites somewhere in the Central American region. Minimum distances to these sources are of the order of 700 km. Crystal sizes in tephra layers at these distances are consistent with their derivation from energetic pyroclastic flow-forming eruptions that injected tephra to stratospheric levels by large-scale co-ignimbrite and plinian-style plumes. Coarsening of the layers during the Miocene peak of explosive volcanism cannot be attributed to any major change in paleowind intensity and is taken to represent the occurrence of more energetic eruptions that were able to disperse tephra over larger areas.