ARTICLE IN PRESS Quaternary Science Reviews 23 (2004) 1333–1357 The periglacial climate andenvironment in northern Eurasia during the Last Glaciation Hans W. Hubbertena,*, Andrei Andreeva, Valery I. Astakhovb, Igor Demidovc, Julian A. Dowdeswelld, Mona Henriksene, Christian Hjortf, Michael Houmark-Nielseng, Martin Jakobssonh, Svetlana Kuzminai, Eiliv Larsenj, Juha Pekka Lunkkak, AstridLys a(j, Jan Mangerude, Per Moller. f, Matti Saarnistol, Lutz Schirrmeistera, Andrei V. Sherm, Christine Siegerta, Martin J. Siegertn, John Inge Svendseno a Alfred Wegener Institute for Polar and Marine Research (AWI), Telegrafenberg A43, Potsdam D-14473, Germany b Geological Faculty, St. Petersburg University, Universitetskaya 7/9, St. Petersburg 199034, Russian Federation c Institute of Geology, Karelian Branch of Russian Academy of Sciences, Pushkinskaya 11, Petrozavodsk 125610, Russian Federation d Scott Polar Research Institute and Department of Geography, University of Cambridge, Cambridge CBZ IER, UK e Department of Earth Science, University of Bergen, Allegt.! 41, Bergen N-5007, Norway f Quaternary Science, Department of Geology, Lund University, Geocenter II, Solvegatan. 12, Lund Sweden g Geological Institute, University of Copenhagen, Øster Voldgade 10, Copenhagen DK-1350, Denmark h Center for Coastal and Ocean Mapping, Chase Ocean Engineering Lab, University of New Hampshire, Durham, NH 03824, USA i Paleontological Institute, RAS, Profsoyuznaya ul., 123, Moscow 117868, Russia j Geological Survey of Norway, PO Box 3006 Lade, Trondheim N-7002, Norway k Institute of Geosciences, University of Oulu, PO Box 3000, Linnanmaa FIN-90014, Finland l Geological Survey of Finland, Betonimiehenkuja 4, PO Box 96, Espoo FI-02151, Uusimaa, Finland m Severtsov Institute of Ecology and Evolution, RAS, Leninskiy Prospect, 33, Moscow 119071, Russia n Bristol Glaciology Centre, School of Geographical Sciences, University of Bristol, Bristol BS8 ISS, UK o Department of Geology, Allegt." 41, Bergen N-5007, Norway Abstract This paper summarizes the results of studies of the Late Weichselian periglacial environments carried out in key areas of northern Eurasia by several QUEEN teams (European Science Foundation (ESF) programme: ‘‘Quaternary Environment of the Eurasian North’’). The palaeoglaciological boundary conditions are defined by geological data on timing and extent of the last glaciation obtained in the course of the EU funded project ‘‘Eurasian Ice Sheets’’. These data prove beyond any doubt, that with the exception of the northwestern fringe of the Taymyr Peninsula, the rest of the Eurasian mainlandandSevernaya Zemlya were not affectedby the Barents–Kara Sea Ice Sheet during the Last Glacial Maximum (LGM). Inversedmodellingbasedon these results shows that a progressive cooling which startedaround30 ka BP, causedice growth in Scandinavia and the northwestern areas of the Barents–Kara Sea shelf, due to a maritime climate with relatively high precipitation along the western flank of the developing ice sheets. In the rest of the Eurasian Arctic extremely low precipitation rates (less than 50 mm yrÀ1), did not allow ice sheet growth in spite of the very cold temperatures. Palaeoclimatic and palaeoenvironmental conditions for the time prior to, during, and after the LGM have been reconstructed for the non-glaciated areas around the LGM ice sheet with the use of faunal and vegetation records, permafrost, eolian sediments, alluvial deposits and other evidences. The changing environment, from interstadial conditions around 30 ka BP to a much colder and drier environment at the culmination of the LGM at 20–15 ka BP, andthe beginning of warming around15 ka BP have been elaboratedfrom the fielddata,which fits well with the modelling results. r 2003 Elsevier Ltd. All rights reserved. 1. Introduction Major progress has been made during recent years in *Corresponding author. Fax: +49-331-2882-137. understanding the timing and extension of the Weichse- E-mail address: [email protected] (H.W. Hubberten). lian glaciations of Eurasia. Basedon comprehensive 0277-3791/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.quascirev.2003.12.012 ARTICLE IN PRESS 1334 H.W. Hubberten et al. / Quaternary Science Reviews 23 (2004) 1333–1357 geological investigations, a new reconstruction of the 2. Results Eurasian Ice Sheets has been presentedfor the Last Glacial Maximum (LGM) at about 18–20,000 years ago 2.1. LGM climate inferred from numerical simulations of (Fig. 1 from Svendsen et al., 2004b). According to this the Eurasian Ice Sheets reconstruction the northwestern fringe of the Taymyr Peninsula was affectedby a restrictedice advance from Siegert et al. (1999, 2001) useda numerical ice-sheet the Kara Sea, but the eastern part of the Eurasian model linked to a simple climate model to reconstruct mainlandandSevernaya Zemlya was not ice-covered the Barents–Kara Sea Ice Sheet as constrainedby the during the LGM. Thus, the ice sheet was much smaller geologically mappedice margins ( Svendsen et al., 1999, during this period than suggested by several previous 2004b). An inverse procedure was employed in which reconstructions, which have shown a large ice mass 2 km the ice-sheet model’s climate inputs (mean-annual thick locatedover most of the Eurasian Arctic (e.g. precipitation andair temperature) were adjusteduntil Denton andHughes, 1981 ; Hughes, 1985; Grosswald a ‘fit’ between the model results and the geological andHughes, 1995, 2002 ; Grosswald, 1998). evidence was established (Siegert andMarsiat, 2001 ). In this paper, we present our current knowledge of the The ice-sheet model is based on the flow law for ice periglacial environment andclimatic conditions in (Glen, 1955). Ice streams were modelled in bathymetric northern Eurasia during the build-up, culmination, troughs by accounting for the reduction in effective ice andwaning of the LGM ice sheets, basedon proxy pressure due to ice buoyancy in water (e.g. Bentley, data from the north of western Russia and Siberia and 1987). Several ice-sheet models based on these principles on glaciological modelling. The geological records are available, andintercomparison experiments indicate indicate that the continental shelves in the Eurasian that such models provide similar results on a continental Arctic were basically free of glaciers at the endof the scale (Huybrechts et al., 1996). Details of the ice-sheet Middle Weichselian (Mangerudet al., 1998 ; Bauch et al., model are provided by Siegert et al. (1999). The model 2001; Kleiber et al., 2001; Knies et al., 2001; Stein et al., requires time-dependent inputs of ice accumulation and 2001), which implies that the Barents–Kara Sea Ice surface air-temperature from a climate model that is Sheet, as it appearedat the LGM, formedduringa linkedto the ice sheet ( Dowdeswell and Siegert, 1999; relatively short periodof time ( Mangerudet al., 1999 ; Siegert et al., 1999). Alexanderson et al., 2001; Gataullin et al., 2001). The The model produces an ice dome 2.5 km thick over eastern flank of the Scandinavian Ice Sheet advanced a Scandinavia, slightly higher than predicted by the ICE- 1000 km during this period (e.g. Lunkka et al., 2001). 4G isostatic model of the same region (Peltier, 1994). This rapidbuild-up of glacier ice represents a response Across the Barents Sea the ice sheet was only 1200 m to a rapidshift in the climate system, which in turn must thick. Fast-flowing ice streams occupying bathymetric also have modified the environment in the ice-free troughs drained ice from the central Barents Sea, and western part of northern Russian andSiberia. this causedthe ice thickness to be generally low ( Fig. 2). A three-dimensional ice-sheet reconstruction for the The entire western margin of the Barents Sea was LGM, forcedto fit the new geological constraints on the glaciatedat the LGM ( Landvik et al., 1998). The eastern ice sheet extent over the Eurasian Arctic, has been margin of the ice sheet was situatedwithin the Kara Sea. generatedusing inverse modelling( Siegert et al., 1999, No glacier was formedon the Taymyr Peninsula, and 2001). From these model experiments, specific climatic there was limitedglacier growth on Severnaya Zemlya parameters (mean-annual precipitation andair tempera- (Fig. 2). The marine margins of the ice sheet lost ice ture) were defined for the land areas along the eastern through iceberg calving, predominantly at the mouths of andsouthern bordersof the ice sheet. The results were bathymetric troughs where the ice streams terminated. subsequently comparedwith geological recordscontain- However, the dominant wastage process along the ing information on climate in the different study areas. southern margin of the ice sheet was surface melting. Both publishedandunpublisheddata from our field The LGM palaeoenvironment impliedby the ice-sheet studies are used to reconstruct the palaeoclimate in the model is characterized by a maritime climate across the areas beyondthe ice fronts ( Fig. 1). Various types of western ice-sheet margin andby extreme polar desert information are usedto characterize the terrestrial conditions towards the east (Fig. 2). Specifically, rates of environment, including faunal and vegetation records, ice accumulation of over 600 mm yrÀ1 are calculated the distribution of permafrost, aeolian sediments and across the western margin of Scandinavia. The mean- alluvial deposits. We have also incorporated results annual surface air-temperature in this region was À5C from some previous projects into this synthesis. Palaeo- (similar