Late Weichselian and Holocene earthquake events along the Geleen fault in NE Belgium: OSL age constraints

Abstract

In 2005, a trench was opened for palaeoseismological analysis near the village of Rotem (Maas valley, NE Belgium). The trench exposed the Geleen fault, a branch of the Feldbiss fault zone of the Lower Rhine graben system. The latter is one of the most active tectonic structures in NW Europe north of the Alps. The sedimentary record in the trench mainly consists of a sequence of fluvio-aeolian and aeolian deposits, initially interpreted as Late Weichselian coversands. A gravel pavement interpreted as a regional deflation horizon subdivides the succession into two main parts, and the uppermost sediments are overprinted by soil formation processes. Detailed mapping of the trench walls revealed evidence for two surface-rupturing earthquakes. We report on the application of quartz-based single-aliquot optical dating to establish a chronological framework for the sediments and to constrain the timing of the two palaeoearthquakes. The optical ages constrain the age of the most recent surface-rupturing event between 2.5±0.3 and 3.1±0.3 ka, a time interval which is independently supported by 14C ages. The penultimate large earthquake occurred just before the formation of the gravel pavement. The age of this pavement, and thus of the second palaeoearthquake, is bracketed between 15.9±1.1 and 18.2±1.3 ka. The optical ages allow correlating the sedimentary succession observed at Rotem with the coversand stratigraphy as established in the southern Netherlands. The age of the pebbly deflation horizon indicates that it formed synchronously with the Beuningen Gravel Bed, while the immediately under- and overlying sediments are correlated with the top of the Older Coversand I and the base of the Older Coversand II units, respectively. The upper part of the coversands on the upthrown side of the fault are dated at ∼6 ka, which is interpreted as an age underestimation owing to bioturbation. The sediments which were originally interpreted as their stratigraphic equivalent on the downthrown side are reliably dated at 6.9±0.7 ka. However, the grain size characteristics of this deposit differ from those of coversands and rather resemble those of overlying colluvium. The origin of this deposit remains enigmatic. It might testify to either an as yet unknown Holocene phase of colluviation, or a third palaeoearthquake for which no evidence was found in the field.