Small Variation in Fluvial Denudation Rates Between Glacial-Interglacial Cycles: Insights from Paleoclimate Modeling and Cosmogenic Nuclides from European River Terraces

Abstract:

Quaternary climate change between glacial-interglacial cycles is commonly thought to induce variations in catchment denudation rates. However, measurements of temporal variations in fluvial denudation are often lacking. Here we present an integration of existing and new cosmogenic nuclide measurements from European river terraces with the output from a (T159, ECHAM5) global atmospheric general circulation model of climate change during glacial and interglacial periods. Cosmogenic nuclide concentrations were measured from river terraces spanning 12 degrees of latitude in unglaciated, tectonically quiescent settings.

Results from new cosmogenic nuclide measurements provide 9 new terrace ages (50 analyses). An additional 25 analyses provides catchment wide paleodenudation rates from terraces. From 0.5-2.0 Ma these data indicate low, and constant (11 +/- 5 mm/kyr) fluvial paleodenudation rates, regardless of the catchment latitude. Modern fluvial denudation rates increase to 20-50 mm/kyr. However, previous higher-fidelity studies of terraces formed since the Last Glacial Maximum (LGM) suggest a factor of 1.5-3 higher denudation rates during the LGM compared to modern. Results from paleoclimate simulations of the LGM, middle Holocene, and modern suggest precipitation rates were 100-500 mm/yr drier than the modern across Europe during glacial periods. Mid-Holocene precipitation rates were ~100 mm/yr drier (SW Europe) to ~200 mm/yr wetter (central Europe) than modern. Predicted LGM temperatures indicate periglacial conditions in some areas. Thus, despite moderate changes in predicted precipitation between glacial and interglacial cycles, there is no clear signal of these cycles in cosmogenic-derived denudation rates between 0.5-2 Ma. In contrast, catchments with higher-fidelity records since the LGM document higher denudation rates during glacial times. We suggest this difference is driven by periglacial, not fluvial, processes. These results have two implications. (1) Variations in temperature and periglacial processes, rather than precipitation modulate catchment denudation rates over glacial interglacial cycles, and (2) The integration time scale of, and uncertainties in, paleodenudation rates between 0.5-2 Ma mask any glacial-interglacial variations in denudation.