Alluvial architecture of the Mississippi valley: predictions using a 3D simulation model

Abstract

Abstract The 3D alluvial-architecture model of Mackey & Bridge (1995, Journal of Sedimentary Research , B65 , 7–31) is used to simulate the nature of avulsion and alluviation of the Mississippi River in its lower alluvial valley and delta plain during the Holocene period, a time of relative sea-level rise. An important objective of these simulations is to examine the controls on the relative proportion of channel-belt deposits and floodplain deposits within the alluvium. Although the model cannot predict the exact avulsion behaviour of the Mississippi, the timing and location of avulsions are predicted well in a general sense. In the lower Mississippi valley, avulsions commonly originate in an up-valley location near Memphis, and the mean avulsion period is 1000–1500 a. This avulsion frequency, combined with large channel-belt width relative to valley width, has resulted in large channel-deposit proportion along the length of the lower alluvial valley. On the delta plain, the model predicts that avulsions are relatively frequent (intervals of decades to centuries) on active deltas where deposition rate on alluvial ridges is greatest and down-valley slopes are least (related to relative sea-level rise). Avulsions originating from locations on the delta plain more distant from the coast occur at longer time intervals (hundreds to thousands of years), because alluvial-ridge topography develops more slowly in these regions. Such avulsions give rise to discrete delta lobes or subdeltas (e.g. Teche, Lafourche). Channel-deposit proportion is generally low on the delta plain, and decreases towards the Gulf of Mexico. This is primarily due to down-valley increase in floodplain width relative to channel-belt width. However, channel-deposit proportion is locally high near avulsion points. These results indicate that change in alluvial architecture during relative sea-level rise in near-coastal settings is more complicated than suggested by recent sequence-stratigraphic models where the proportion of channel-belt deposits is postulated to be controlled primarily by only two factors: deposition rate and valley width.