Dates and simulations of fluvial terrace development and measurements of rock erosion rates
R. S. Anderson
University of California, Santa Cruz
1998
192
Ph.D.
University of California, Santa Cruz
1998
Fluvial strath terraces provide a record of river incision and the timing of climatic perturbations to the fluvial system. We measure Be and Al profiles to date terraces along the Wind River, Wyoming, with correction for nuclide inheritance. These terraces appear to reflect oscillation in headwater glaciation. Dates from one terrace (WR-3) are 118-125 ka, which agree with dates on moraines, and are consistent with the model of terrace-glacial relationship. Nuclide inheritance is significant in clasts. We obtain mean rates of exhumation of 13-130 m/My at the source and usd{>}{\sim}usd10 ka travel times using this inheritance. Our older terrace ages appear too young; a plausible explanation is eolian deflation. Our efforts emphasize that geologic caution, independent ages, and multiple sites are needed to date surfaces with cosmogenic radionuclides. We use a one-dimensional numerical stream simulation model incorporating sediment transport and vertical and lateral rock erosion to explore stream response to variations sediment supply, sediment size, and water discharge. Model results suggest that in order to form terrace sequences: (1) input variability must be sufficient to produce high amplitude variations in vertical erosion; and (2) rapid valley widening rates to generate sediment accommodation space. Terrace formation timing is dependent on the dominant input and lags the input cycle. The formation of terrace floors may span thousands of years. Observations on the Indus River, Pakistan, suggest that quarrying and abrasion are the primary rock erosion processes. Quarrying is the most efficient process when joints are close, but requires "preparation" to free a block. The block thickness a river can quarry increases with usd\nu\sp2.usd Rock abrasion should increase with usd{\sim}\nu\sp5,usd and is most effective in separated flow, suggesting it occurs primarily by suspended sediment. Calculations suggest cavitation is not a major process. Total abrasion measured over 1 year using drill holes is usd\leusd4 mm, and cosmogenic radionuclides from the same bed reveal average erosion rates of 0.25-0.50 mm/yr over 1.5-2.0 ka. We reconcile these measurements by appealing to the migration of bedforms. Incision rates are controlled by conditions below the resolution of reach-based erosion rules, presenting a significant modeling challenge.