Hillslope erosion and weathering rates in Earth's most rapidly uplifting mountains.

摘要

Knowledge of hillslope erosion rates and processes is necessary for understanding landscape response to tectonic and climatic forcing and for determining the degree to which mountains regulate biogeochemical cycles and global climate. Landslide erosion and soil production are the principle denudation processes in high-relief terrain, but quantitative estimates of landslide erosion on spatial and temporal scales relevant to landscape evolution are lacking, and there have been no prior measurements of soil production and weathering rates in Earth's most tectonically-active landscapes. Here, I address both of these problems, first by attempting to overcome the inherent difficulty in quantifying landslide erosion rates using a compilation of geometry measurements from 4,231 landslides. I use the geometry to develop scaling relationships that can be used to predict landslide volume from more readily available landslide area information. A key finding is that landslide scaling is controlled by hillslope material; soil landslides have lower depths and hence lower power-law volume-area scaling exponents than bedrock landslides, which has significant implications for accurately quantifying landslide erosion rates. By applying the landslide volume-area scaling relationship to over 15,000 landslides in the Tsangpo Gorge region of the eastern Himalaya, I demonstrate that landslide erosion rates are spatially coupled with stream power and long-term exhumation rates, but become decoupled from hillslope gradients when hillslope angles exceed 30°. These results indicate landslide erosion is coupled with bedrock river incision and rock uplift, but not topography, hence providing the first direct confirmation of a 'threshold hillslope' model of landscape evolution that has emerged over the last two decades. I address the role soils play in the denudation of rapidly uplifting mountains by developing soil production rate and catchment scale denudation data for the western Southern Alps of New Zealand. Soil production rates in the western Southern Alps can exceed those measured elsewhere by more than an order of magnitude and soil physical erosion rates are linearly coupled with chemical weathering rates. Using the relationship between physical and chemical denudation rates to model global weathering fluxes as a function of mean local slope, I demonstrate that the small, mountainous fraction of Earth's surface dominates the global chemical weathering flux. The weathering measurements and model results hence overturn the view that there are 'speed-limits' to soil production and that erosion and weathering are decoupled in mountains, and instead strongly support the hypothesis that mountain uplift influences global climate over geological timescales via links among topography, erosion, weathering, and CO2 cycling.