How does a single precipitation event erode a landscape? Clues from meteoric 7Be and 10Be analysis of suspended sediments and soils
Abstract
Stream sediment contains a history recorded in isotopes that cling to suspended particles. In this study we exploit this recorded history in order to understand how a single precipitation event erodes the landscape at two watershed sites (Bisley I and Mameyes) within the Luquillo Critical Zone Observatory, Puerto Rico. We use fallout cosmogenic radionuclides Beryllium-7 (7Be) and Beryllium-10 (10Be) to determine the provenance of suspended sediment at various stages of a hydrograph. Sediments from source areas within the watersheds, such as stable ridge crests and active landslide scars, were also sampled and analyzed. Exploiting the large difference in half-life, the 10Be/7Be ratio of suspended sediments coupled with the concentration and nature of organic material present show original depth of mobilized stream sediment in the hillslope. The storm hydrographs of a one-month recurrence interval storm on June 7th, 2011 were sampled at both watersheds. In the small watershed (0.067 km2), storm discharge and total suspended solids (TSS) show short lag times between the initiation of precipitation and the initial rise of the hydrograph and no lag time between peak discharge and peak TSS. The larger site (17.8 km2) had a lag time of approximately 30 minutes between the initiation of precipitation and a rise in discharge and had a 15-minute lag between peak stage (which occurred first) and peak TSS, highlighting the longer travel distances that particles must take to reach the stream sampling point in the larger basin. We compare fallout 7Be nuclide concentration in source sediments and assume a simple, two end-member model to mix these sources in the stream. Soil sediments collected from stable ridge crests ('old') have relatively high average 7Be concentrations of 2.7x106 atoms/g±10% and sediments collected from active landslide scars ('new') have relatively low 7Be concentrations of 4.0x104 atoms/g±15%. Suspended sediments had an average 7Be concentration of 7.2x105 atoms/g±8%, indicating a mixing of sediment pools. Results from the two-component mixing-model indicate that, in the small watershed, new sediments accounted for 84% of newly mobilized total suspended sediments while old sediments only accounted for 16% as a discharge weighted mean over the entire hydrograph. The highest peak of old sediment (24% of TSS) at this site corresponds to the peak in stage, with old sediment percentages decreasing from peak stage to the end of sampling. Peak new sediment occurred at the rising limb (81% of TSS) and immediately after the receding limb (84% of TSS) of the hydrograph. These results are consistent with the fact that landslide scars mar the small watershed and expose bare, non-vegetated, deep-sourced material for transport off of the hillslopes during storm events. In the large watershed, the storm mobilized more stable soils (36%) and less new sediment (64%) than in the small watershed. Peak old sediment is witnessed after the peak stage within the larger watershed site, where it stabilizes around 50% for approximately an hour, then decreases with the receding limb of the hydrograph. Conversely, peak new sediment occurred during the rising limb, then decreased during peak flow, then began to increase during the later portion of the receding limb.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- Bibcode:
- 2012AGUFM.V13G..03O
- Keywords:
-
- 1150 GEOCHRONOLOGY / Cosmogenic-nuclide exposure dating;
- 1625 GLOBAL CHANGE / Geomorphology and weathering;
- 1825 HYDROLOGY / Geomorphology: fluvial