Quantification of Quaternary loess cover and integrated denudation rates using cosmogenic nuclide Aluminium-26 and Beryllium-10 disequilibrium (Mecsek Mountains, Pannonian Basin)

In areas of multiple exposure-burial histories the use of two cosmogenic radionuclides (CRN) with different half-lives allows to reveal the disequilibrium between CRN concentrations and provide a better understanding of landscape evolution. This study aims to quantify bedrock denudation rates in the Western Mecsek Mts (southern Pannonian Basin), a low-elevation hilly area that is currently being exhumed from under its loess cover deposited during the Quaternary glaciations. Concentrations of ¹⁰Be and ²⁶Al were measured in samples taken from flat, soil-covered ridge tops and stream sediments from several small river catchments composed of Permian and Triassic sandstones and conglomerates. Low ²⁶Al/¹⁰Be ratios are indicative of unsteadiness caused by significant past burial of the bedrock surfaces. A Monte Carlo (MC) model was developed to reveal the temporal evolution of the loess cover as a function of glacial-interglacial climate, and to determine the true rate of bedrock denudation accounting for both loess-covered periods with shielding and zero erosion as well as phases of exposure and bedrock denudation during periods without loess.

Our model showed that higher-elevation catchments and ridges were exposed for 40 to 85%, while lower-elevation areas were uncovered for less than 30% of time during the last 1 Ma. The modelled time integrated bedrock denudation rates were similar for the ridge crests and basin-averaged samples suggesting a steady relief. However, a well-expressed difference was found between the areas spending most of the time loess covered and the less covered group with mean integrated denudation rates of 5±5 m/Ma and 19±8 m/Ma, respectively. Single nuclide ¹⁰Be denudation rates overestimated the modelled, time-integrated denudation rates by a factor of ~1.5 for the more exposed group and by a factor of ~13 (up to >30) for the mostly covered areas. These rates, especially the latter, are slower than published values for similar climatic, tectonic, and topographic settings. If the simple, single nuclide ¹⁰Be approach was used, these differences would have remained hidden, and the true lowering rate of bedrock would have been overestimated by a factor that increases with the shielding time.

This is the first study quantifying the influence of past loess covers on CRN concentrations in bedrock and to estimate the denudation rates corrected for this shielding. Our findings reveal that the steady-state assumption of the CRN concentrations may also be violated in small, non-glaciated catchments without intermittent sediment storage. Where single-nuclide ¹⁰Be denudation rates are higher than sediment-trap estimates, the shielding effect of past sediment cover (such as loess) could also explain the discrepancy. Accordingly, we recommend the use of the paired ²⁶Al/¹⁰Be approach to test the presumption of cosmogenic nuclide equilibrium not only in large catchments and formerly glaciated areas, but also in settings where past sediment cover may have lasted long enough to lower the CRN ratio.

Funding: PURAM, Mecsekérc Ltd., NKFIH project FK 124807. Sample processing: Cosmogenic Laboratories of Budapest (n=16) and of the University of Edinburgh (n=4); AMS measurements: ASTER, Aix en Provence (n=16) and SUERC, Glasgow (n=4)