(MIS3 & 2) millennial oscillations in Greenland dust and Eurasian aeolian records e A paleosol perspective Denis-Didier Rousseau a, b, * , Niklas Boers a , Adriana Sima a , Anders Svensson c , Matthias Bigler d , France Lagroix e , Samuel Taylor e , Pierre Antoine f a Ecole Normale Superieure, UMR CNRS 8539, Laboratoire de Meteorologie Dynamique, and CERES-ERTI, 24 rue Lhomond, 75231 Paris Cedex 5, France b Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY 10964, USA c Centre for Ice and Climate Niels Bohr Institute, University of Copenhagen, Juliane Maries Vej 30, DK-2100 Copenhagen OE, Denmark d University of Bern, Physics Institute, Climate and Environment Physics, Sidlerstrasse 5, CH-3012 Bern, Switzerland e Institut de Physique du Globe, UMR CNRS 7154, 1 rue Jussieu, 75005, Paris, France f CNRS, Laboratoire de Geographie Physique: Environnements Quaternaires et Actuels, UMR CNRS 8591,1 Place Aristide Briand, 92195 Meudon, France article info Article history: Received 6 October 2016 Received in revised form 9 May 2017 Accepted 15 May 2017 abstract Since their discovery, the abrupt climate changes that punctuated the last glacial period (~110.6e14.62 ka) have attracted considerable attention. Originating in the North-Atlantic area, these abrupt changes have been recorded in ice, marine and terrestrial records all over the world, but especially in the Northern Hemisphere, with various environmental implications. Ice-core records of unprecedented temporal resolution from northern Greenland allow to specify the timing of these abrupt changes, which are associated with sudden temperature increases in Greenland over a few decades, very precisely. The continental records have, so far, been mainly interpreted in terms of temperature, precipitation or vegetation changes between the relatively warm Greenland Interstadials(GI) and the cooler Greenland Stadials(GS). Here we compare records from Greenland ice and northwestern European eolian deposits in order to establish a link between GI and the soil development in European mid- latitudes, as recorded in loess sequences. For the different types of observed paleosols, we use the correlation with the Greenland records to propose estimates of the maximum time lapses needed to achieve the different degrees of maturation and development. To identify these time lapses more pre- cisely, we compare two independent ice-core records: d 18 O and dust concentration, indicating variations of atmospheric temperature and dustiness in the Greenland area, respectively. Our method slightly differs from the denition of a GI event duration applied in other studies, where the sharp end of the d 18 O decrease alone denes the end of a GI. We apply the same methodology to both records (i.e., the GIs are dened to last from the beginning of the abrupt d 18 O increase or dust concentration decrease until the time when d 18 O or dust recur to their initial value before the GI onset), determined both visually and algorithmically, and compare them to published estimates of GI timing and duration. The duration of the GI and consequently the maximum time for paleosol development varies between 200 and 4200 years when visually determined and between 200 and 4800 years when estimated algorithmically for GI 17 to 2, i.e. an interval running from 60 ka to 23 ka b2k (age before 2000 AD). Furthermore, we investigate the abruptness of the transition from stadial to interstadial conditions, which initiates the paleosol devel- opment. The average transition duration is 55.4 ± 16.1 (56.8 ± 19.6) years when determined visually, and 36.4 ± 13.4 (60.00 ± 21.2) years when determined algorithmically for the d 18 O (dust concentration). The d 18 O increases correspond to a mean temperature difference of 11.8 C on the top of the Greenland ice sheet, associated with substantial reorganizations of the ecosystems in mid-latitude Europe. © 2017 Elsevier Ltd. All rights reserved. 1. Introduction The last glacial period (110.6e14.6 ka) has been punctuated by strong and abrupt climate variations at millennial timescales, rst * Corresponding author. Ecole Normale Superieure, UMR CNRS 8539, Laboratoire de Meteorologie Dynamique, and CERES-ERTI, 24 rue Lhomond, 75231 Paris Cedex 5, France. E-mail address: denis.rousseau@lmd.ens.fr (D.-D. Rousseau). Contents lists available at ScienceDirect Quaternary Science Reviews journal homepage: www.elsevier.com/locate/quascirev http://dx.doi.org/10.1016/j.quascirev.2017.05.020 0277-3791/© 2017 Elsevier Ltd. All rights reserved. Quaternary Science Reviews 169 (2017) 99e113