Paleoclimate variability during the Blake geomagnetic excursion (MIS 5d) deduced from a speleothem record

Abstract

To evaluate possible connections between climate and the Earth's magnetic field, we examine paleoclimate proxies in a stalagmite (PA-8) recording the Blake excursion (∼112–∼116.4 ka) from Cobre cave (N Spain). Trace element, δ13C, δ18O, δ234U, fluorescent lamination, growth rate, and paleomagnetic records were synchronized using a floating lamina-counted chronology constrained by U–Th dates, providing a high-resolution multi-proxy paleoclimate record for MIS 5d. The alpine cave setting and the combination of proxies contributed to improve the confidence of the paleoclimatic interpretation. Periods of relatively warm and humid climate likely favored forest development and resulted in high speleothem growth rates, arguably annual fluorescent laminae, low δ13C and [Mg], and increased [Sr] and [Ba]. Colder periods limited soil activity and drip water availability, leading to reduced speleothem growth, poor development of fluorescent lamination, enhanced water–rock interaction leading to increased [Mg], δ13C, and δ234U, and episodic flooding. In the coldest and driest period recorded, evaporation caused simultaneous 18O and 13C enrichments and perturbed the trace element patterns. The Blake took place in a relatively warm interestadial at the inception of the Last Glacial period, but during a global cooling trend recorded in PA-8 by an overall decrease of δ18O and growth rate and increasing [Mg]. That trend culminated in the cessation of growth between ∼112 and ∼101 ka likely due to the onset of local glaciation correlated with Greenland stadial 25. That trend is consistent with a link between low geomagnetic intensity and climate cooling, but it does not prove it. Shorter term changes in relative paleointensity (RPI) relate to climate changes recorded in PA-8, particularly a prominent RPI low from ∼114.5 to ∼113 ka coincident with a significant cooling indicated by all proxy records, suggesting a link between geomagnetic intensity and climate at millennial time scales. Although the reliability of such inference is limited by the inferior resolution of the paleomagnetic data and a possible contamination of the RPI data by unaccounted changes in magnetite concentration, the agreement of those data with the marine record of cosmogenic Be suggests that the RPI record of PA-8 may reflect truly geomagnetic intensity variations.