Controls on Carbonate Hillslope Morphology in sub-Humid Climate

The convexity of soil-covered hillslopes has long been recognized to reflect transport mechanisms of regolith. Intense previous research using cosmogenic isotopes and numerical models mostly focused on the influence of tectonics and erosion rates in regions with silicate rocks, generally downplaying the influence of parameters such as climate, chemical denudation rates and dust flux. The goal of the current research is to characterize and understand how hillslope curvature in carbonate rocks (∂2z/∂x2) vary as a function of precipitation and dust flux. The study area is located in the Eastern Mediterranean across a prominent north-to-south gradient in both precipitation (300 to 1000 mm yr-1) and dust flux (150 to 40 g m-2 yr-1). We combine topographic analysis using high resolution data from airborne LiDAR, field measurements, and numerical modeling. The soil creep diffusion coefficient is evaluated using two different methods: (1) Calculated directly at sites with known cosmogenically-derived erosion rate; (2) Estimated using numerical models at snmites with known initial hillslope form at a known point in time such as volcanic cinder cones with known eruption ages. Initial results indicate that soil-mantled, carbonate hillslopes in the sub-humid Eastern Mediterranean are convex and parabolic in form adjacent to the divide and hillslope curvature is strongly correlated with precipitation in an inverse manner. This dependency could reflect: (1) soil creep diffusion coefficients which increase with precipitation, (2) chemical denudation whose relative fraction of the total denudation increases with precipitation, and (3) dust flux which decreases with precipitation. The relative contribution of these 3 components is currently being evaluated. Diffusion coefficients were estimated based on numerical modeling of high-resolution, field-measured, hillslope profiles of several cinder cones, across a rainfall gradient from 600 to 900 mm yr-1, without apparent variability in the dust flux. Currently obtained values vary from ~10-2 to ~10-3 m2 yr-1 with no clear-cut precipitation dependency.