Validating a Swedish varve chronology using radiocarbon, palaeomagnetic secular variation, lead pollution history and statistical correlation

Abstract

We use statistical correlation of palaeomagnetic secular variation (PSV) curves from a varved Holocene lake sediment sequence in west central Sweden (Lake Kälksjön) against those of a Fennoscandian master stack (FENNOSTACK) to correct for an apparent error in the varve chronology. Additional correlation between a lead pollution-derived chronology for the last 2000 years corroborates the PSV results. Use of the FENNOSTACK palaeomagnetic master curve reveals no significant difference in duration between large-scale features from ∼2500 to ∼8000 cal. yrs BP. Statistical correlation, however, implies that 270 years are missing from the younger part (<1000 cal. yrs BP) of the varve chronology, and that there is an overestimation by approximately 230 years in the number of varves counted in the early Holocene (>8000 cal. yrs BP). A similar comparison between the PSV-determined ages and calibrated bulk radiocarbon ages suggests that the sediments of mid-Holocene age contain substantial amounts of old carbon, probably of soil origin, which causes bulk sediment-calibrated mean 14C ages to be up to 850 years older than the corrected varve chronology, which extends to 9193 ± 186 cal. yrs BP. This study highlights both the use of statistical correlation as a technique for detecting errors between chronologies, and the importance of validating incremental chronologies with more than one independent method.