UNDERSTANDING MODERN LANDSCAPE BEHAVIOR USING METEORIC AND IN SITU 10BE AND 137CS IN LARGE RIVER BASINS, SW CHINA

We measured ¹³⁷Cs and meteoric ¹⁰Be in 35 samples previously measured for in situ ¹⁰Be (Henck et al., 2011) in the Salween, Mekong, and Yangtze river systems in SW China to better understand long-term landscape behavior and effects of human-induced landscape change. Samples were collected from basins spanning ~10 to 10⁵ km². Meteoric ¹⁰Be concentration varied by an order of magnitude, from 3.4 to 76 x 10⁶ atoms/g. There was no systematic spatial patterning in meteoric ¹⁰Be values. ¹³⁷Cs activity was only measureable in five samples.

No significant relationships were found between meteoric ¹⁰Be concentrations and mean local relief, mean annual rainfall, or basin area. Areas with high erosion rates (>0.4 mm/yr) had low meteoric ¹⁰Be values (R²= 0.34, p= 0.0002). While meteoric ¹⁰Be and basin size did not strongly correlate, basins <10⁴ km² had highly variable meteoric ¹⁰Be values, while basins >10⁴ km² had significantly less variability (p= 0.11). Analysis of meteoric ¹⁰Be, mean annual rainfall, and mean local relief grouped by stream order suggests high variability in small basins may mask relationships found in large basins.

Erosion indices (EI) were calculated for each sample using in situ ¹⁰Be erosion rates to calculate long-term sediment flux and ranged from 0.11 to 1.48 (median = 0.55). The EI represents the ratio of total meteoric ¹⁰Be leaving the basin on sediment grains over the total estimated atmospheric delivery of ¹⁰Be. The highest EI’s were in the northern portion of the Yangtze drainage, and corresponded to moderate erosion rates of 0.06 – 0.13 mm/yr. Variability in EI was significantly higher in basins <10⁴ km² (p= 0.14).

EI’s calculated from long-term sediment yield allow modern events, such as land-use change, to be isolated from long-term geomorphic trends. A median EI of 0.55, with 91% of measurements below 1, suggests that less meteoric ¹⁰Be is exported than is incident upon the landscape. The absence of measureable ¹³⁷Cs in the majority of samples indicates that significant erosion has occurred since 1954. EI and ¹³⁷Cs activity suggest that there is a “disconnect” between long-term sediment yield estimates and modern erosion rates, potentially caused by human impact changing the style and/or distribution of erosion.