POINT BAR AND OVERBANK DEPOSIT STORAGE TIME DISTRIBUTIONS IN A SIMULATED MEANDERING RIVER

Floodplain storage durations are important geologically and for managing pollutants. They provide necessary input data for sediment routing models, but the functions that describe storage are hard to define with field data. Using a meandering river model, a 36ka simulation is analyzed in 20 year time steps. Channel length, and eroded, deposited, and stored sediment volumes stabilize in under 2ka while the region explored by the river expands as time^0.3. The simulated channel was 70m (3 pixels) wide and freely migrated within a 30 by 40km reach simulated with constant erodibility, and mean migration rate ±0.06 channel widths per year. No incision or aggradation was permitted. Sediment storage was partitioned by elevation. Point bar deposits are emplaced immediately as the channel moves away and comprise the lower 40% of a vertical section, while the upper 60% of the floodplain consists of vertically accreted ‘overbank’ deposits that accumulate with decreasing rates as elevation and distance from the channel increase. Point bar deposits have the highest probability of erosion at 80 years, and contribute 2/3 of eroded material despite representing only 40% of the floodplain volume. When eroded, vertical accretion deposits are 1.25 times older than point bar deposits. Apart from these differences, the storage distributions for both deposits are very similar, though there is much more scatter in the point bar distributions. Eroded sediment <2ka in age are well fit with exponential functions, and the remaining material (approximately 20%) >2ka in age is well fit with a power law distribution. On average, 10% of eroded material is <80 years old, 30% from 80 – 500 years, 38% from 500 – 2000 years, 15% from 2ka to 10ka and 7% of eroded material is >10ka in age. However, the fraction older than 10ka is widely variable (0% to 30%, σ = 6.8%) in any 20 year interval. The patchiness and variability in the age of eroded sediment observed in these model results may be similar to natural settings and would be difficult to quantify or address with field measurements.