[1] Vertical knickpoints (waterfalls) mark a prominent process transition zone whose governing mechanics is not represented by conventional stream power incision models. We examine the evolution of vertical knickpoints with resistant caprock utilizing numerical simulations that explicitly represent (1) face failure mechanisms, (2) flow acceleration and amplified erosion above a knickpoint lip, (3) deposition and removal of coarse debris below the knickpoint, and (4) base level lowering or tectonic uplift rates. Our model demonstrates that knickpoint retreat rate, where the subcaprock is weak or vertically jointed and base level fall rates are steady, is likely to become tied to downstream conditions and equal to the downstream incision rate divided by channel gradient. Mechanically, this coupling occurs where the subcaprock reaches a threshold height for failure in shear or buckling or where the weathering rate of the subcaprock is higher than the downstream incision wave velocity (V i_ds ). The height of the subcaprock face can influence its gravitational stability and the knickpoint lateral erosion rate and lead to a feedback between downstream incision and retreat rate. Retreat rate can be lower than V i_ds during transients, which could be long (>10 6 years) and set by the weathering rate of the subcaprock or influenced by lag debris evacuation. Key variables other than discharge can be important in setting retreat rates. These include base level lowering rate, the rock strength of stratigraphic units downstream of the knickpoint, and the size and flux of sediment contributed from above the knickpoint or from the canyon walls. Two types of oversteepened reaches can form in association with a vertical knickpoint: (1) an upstream, free fall-induced, oversteepened reach whose length is longer than the flow acceleration zone and (2) a downstream coarse debris-induced oversteepened reach. Although the model was constructed with caprock-type knickpoints in mind, some of its elements and insights are also relevant to homogenous substrates.

No abstract

We measured 10 Be in fluvial sediment samples (n ‫؍‬ 27) from eight Great Smoky Mountain drainages (1-330 km 2). Results suggest spatially homogeneous sediment generation (on the 10 4-10 5 yr time scale and Ͼ100 km 2 spatial scale) at 73 ؎ 11 t km ؊2 yr ؊1 , equivalent to 27 ؎ 4 m/m.y. of bedrock erosion. This rate is consistent with rates derived from fission-track, long-term sediment budget, and sediment yield data, all of which indicate that the Great Smoky Mountains and the southern Appalachians eroded during the Mesozoic and Cenozoic at ϳ30 m/m.y. In contrast, unroofing rates during the Paleozoic orogenic events that formed the Appalachian Mountains were higher (Ն10 2 m/m.y.). Erosion rates decreased after termination of tectonically driven uplift, enabling the survival of this ancient mountain belt with its deep crustal root as an isostatically maintained feature in the contemporary landscape.

No abstract

All exposed rocks on Earth's surface experience erosion; the fastest rates are documented in rapidly uplifted monsoonal mountain ranges, and the slowest occur in extreme cold or warm deserts-millennial submeterscale erosion may be approached only in the latter. The oldest previously reported exposure ages are from boulders and clasts of resistant lithologies lying at the surface, and the slowest reported erosion rates are derived from bedrock outcrops or boulders that erode more slowly than their surroundings; thus, these oldest reported ages and slowest erosion rates relate to outstanding features in the landscape, while the surrounding landscape may erode faster and be younger. We present erosion rate and exposure age data from the Paran Plains, a typical environment in the Near East where vast abandoned alluvial sur-faces (10 2 -10 4 km 2 ) are covered by well-developed desert pavements. These surfaces may experience erosion rates that are slower than those documented elsewhere on our planet and can retain their original geometry for more than 2 m.y. Major factors that reduce erosion converge in these regions: extreme hyperaridity, tectonic stability, fl at and horizontal surfaces (i.e., no relief), and effective surface armoring by a clast mosaic of highly resistant lithology. The 10 Be concentrations in amalgamated desert pavement chert clasts collected from abandoned alluvial surfaces in the southern Negev, Israel (representing the Sahara-Arabia Deserts), indicate simple exposure ages of 1.5-1.8 Ma or correspond to maximum erosion rates of 0.25-0.3 m m.y. -1 . The 36 Cl in carbonate clasts, from the same pavement, weathers faster than the chert and yields simple exposure ages of 430-490 ka or maximum erosion rates of 0.7-0.8 m m.y. -1 . These ages and rates are exceptional because they represent an extensive landform. The 10 Be concentrations from samples collected at depth and optically stimulated luminescence (OSL) dating reveal a two-stage colluvial deposition history followed by eolian addition of 40 cm of silt during the past 170 k.y. Our results highlight the effi ciency of desert pavement armor in protecting rocks from erosion and preserving such geomorphic surfaces for millions of years.