Numerical Modeling of Mud Transport, Storage and Release On The Colorado River, Arizona

The construction and operation of Glen Canyon Dam on the Colorado River has greatly affected downstream hydrology, sediment transport, and geomorphology, thereby impacting the aquatic and riparian ecosystems. Predictive models are currently used to estimate the accumulation and evacuation of sand and to design controlled floods to redistribute sand from the riverbed and rebuild sandbars. Although sand is the primary component of most fine-sediment deposits, there is increasing interest in mud (silt and clay) transport in Grand Canyon owing to its effect on turbidity, which has been shown to affect native and non-native fish and control gross primary production. Mud is predominantly supplied by tributary floods. While most of the mud washes downstream, some mud is trapped within the bed and in bar deposits and potentially released later, increasing turbidity. Here we present preliminary results from a one-dimensional model being developed for predicting fine-sediment concentrations (sand, silt, and clay) in the Colorado River downstream from Glen Canyon Dam. The model includes sediment advection, eddy exchange, exchange with the bed, and storage/release from bar deposits. Comparison between model results and 15-minute acoustical measurements of silt-and-clay concentration indicate that including exchange between the main channel and lateral flow recirculation zones (eddies) is essential for accurately modeling the timing and attenuation of mud pulses as they travel downstream. Sand-concentration measurements made following tributary floods show that the finest sand fractions supplied during these floods travel only slightly slower than the tributary-generated mud pulse in the Colorado River for hundreds of kilometers, despite estimated advection lengths (characteristic distance a particle travels in suspension before exchanging with the bed) of less than a kilometer. This suggests that exchange of sand between the flow and the bed is facilitated by a thin bed surface layer immediately following tributary floods, with deeper bed mixing occurring over longer time-scales due to bedform migration. Finally, analysis of silt/clay measurements and model results indicates that a small percentage of the mud from tributary pulses is trapped within bed and bar deposits and gradually released. Discharges above 15,000 cfs are associated with faster release of silt/clay from bed and bar deposits, but we find that some mud remains trapped within bar deposits even after high flows. Our results indicate that while mud concentrations are primarily controlled by summer-thunderstorm and winter tributary inputs, which reset the system by reloading bed and bar deposits with mud, discharge plays a secondary but important role in regulating mud concentrations and turbidity in the Colorado River.