Comparison of High-Resolution 14C and Luminescence-Based Chronologies of the MIS 2 Madaras Loess/Paleosol Sequence, Hungary: Implications for Chronological Studies

Abstract

Numerous loess/paleosol sequences (LPS) in the Carpathian Basin span the period of Marine Isotope Stage (MIS) 2 and the last glacial maximum (LGM). Nevertheless, only two known records—Madaras and Dunaszekcső—preserve highly resolved records with absolute chronologies with minimal uncertainties, which enable the meaningful assessment of feedbacks and short-term climatic fluctuations over this period. The Madaras profile is located at the northern margin fringe of the Bácska loess plateau; Dunaszekcső, located on the Danube to its west, yields a chronology built on over 100 14C dates yet spans only part of MIS 2, missing half of the LGM including its peak. Here, we add to the previously published 14C chronology for Madaras (15 dates) with an additional 17 14C and luminescence ages. Resulting age models built solely on quartz OSL and feldspar pIRIRSL data underestimate the 14C based chronology, which is likely based on inaccuracies related to luminescence signal behavior; we observe age underestimations associated with unusual quartz behavior and significant signal loss, a phenomenon also observed in Serbian and Romanian loess, which may relate to non-sensitized grains from proximal sources. Our new chronology provides higher resolution than hitherto possible, yielding consistent 2 sigma uncertainties of ~150–200 years throughout the entire sequence. Our study indicates that the addition of further dates may not increase the chronological precision significantly. Additionally, the new age model is suitable for tackling centennial-scale changes. The mean sedimentation rate based on our new age-depth model (10.78 ± 2.34 years/cm) is the highest yet recorded in the Carpathian Basin for MIS 2. The resolution of our age model is higher than that for the Greenland NGRIP ice core record. The referred horizons in our profile are all characterized by a drop in accumulation and a higher sand input, the latter most likely deriving from nearby re-exposed sand dunes.