DISENTANGLING THE EFFECTS OF CLIMATE, ROCK STRENGTH, AND TECTONICS ALONG THE CASCADIA FOREARC (Invited Presentation)

In the three decades since Merritts and Ellis (1994) galvanized a community of researchers exploring the interplay between geodynamics and surface processes, the field of tectonic geomorphology has made significant progress in quantitative reconstruction of uplift rates and patterns from landscape morphology. However, challenges remain in ascertaining how climate and discharge variability conspire with rock strength, sediment delivery, and grain size to modulate channel profile steepness and landscape relief. Here, we exploit a densely sampled data set erosion rate quantified using ¹⁰Be concentrations in fluvial sediment along with analysis of channel steepness from ~90 watersheds spanning the length and width of the Cascadia forearc. We evaluate the competing roles of rock mass quality, variations in precipitation and runoff, and differential rock uplift in setting topography in the overriding plate of the Cascadia subduction zone. Our results reveal that erosion rates across the Siletzia terrane in Washington and Oregon vary from ~60–120 m/Myr, with elevated rates (up to ~200 m/Myr) in the Tillamook Highlands. Erosion rates increase abruptly near ~43°N to ~300 m/Myr in watersheds draining the western Klamath and Siskiyou Mountains. These elevated erosion rates decrease away from the plate boundary to ~50 m/Myr on the eastern flank of the forearc. Watersheds span a range of mean annual precipitation from ~0.7–4 m/yr, and variations in runoff appear to systematically influence the relationship between channel steepness and erosion rate. Calibration of a stochastic-threshold stream power model against local runoff records spanning ~40 years provides a rational framework in which to interpret the adjustment of channel profile form and erosion rate. Predicted erosion rates from this calibrated model compare favorably to long-term fluvial incision rates determined from dated Late Pleistocene strath terraces in the Illinois and Rogue rivers. Thus, our results suggest that carefully calibrated stochastic-threshold stream power models have the potential to provide sufficient resolving power to model the erosional response of landscapes to tectonic forcing over 10⁵ – 10⁶ yr.