Evaluation of a Laser-Acoustic System for Continuously Monitoring Suspended-Sediment Concentration and Grain Size in the Colorado River in Grand Canyon

Sandbars and other sandy deposits in and along the Colorado River in Grand Canyon National Park (GCNP) were an integral part of the pre-dam riverscape, and are important for habitat, protecting archeological sites, and recreation. These deposits have eroded substantially following the 1963 closure of Glen Canyon Dam that reduced the supply of sand at the upstream boundary of GCNP by about 94%; sandbars in the upstream portion of Grand Canyon have decreased in size by about 25% during only the last 15 years. Recent work has shown that sand transport in the post-dam river is supply limited, and is equally regulated by the discharge of water and short-term changes in the grain size of sand available for transport. During and following tributary floods, fine sand supplied to the Colorado River travels downstream as an elongating sediment wave. As the front of a sediment wave passes a given location, sand on the bed first fines and sand-transport rates increase independently of the discharge of water. Subsequently, the bed is winnowed and sand-transport rates decrease independently of discharge. By virtue of this process, sand supplied by tributaries is typically exported from the upstream portion of Grand Canyon within months under normal dam releases. Thus, newly input sand may be available to rebuild sandbars during controlled floods conducted only following large tributary floods. Accurate monitoring of sand transport in such a river requires frequent measurements of suspended-sediment concentration and grain size, and cannot be accomplished by using stable sediment-rating curves constructed from a sparser dataset of suspended-sediment measurements. To monitor sediment transport in the Colorado River, we have designed and are evaluating a laser-acoustic system for measuring the concentration and grain size of suspended sediment every 15 minutes. This system consists of (1) a subaqueously deployed laser-diffraction instrument (either a LISST 100 or a LISST 25X) connected to an automatic pump sampler, and (2) an acoustic-doppler current meter. When laser transmission drops below a user-defined threshold (as a result of increased suspended-sediment concentrations), the LISST triggers the automatic pump sampler to collect samples at a user-defined rate. This allows samples to be collected when the suspended-sediment concentrations exceed the upper limit for the LISST and the acoustic-doppler current meter (around 2000-3000 mg/l). Beginning in August 2002, we began testing this system at 4 locations along the Colorado River in Grand Canyon, and have developed stable box coefficients relating the pump, laser-diffraction, and acoustic-backscatter point measurements to cross-sectionally integrated measurements of suspended-sediment concentration and grain size. Variability between either sequential laser-diffraction or acoustic-backscatter measurements is substantially less than the variability between sequential cross-sectionally integrated measurements of concentration and grain size (collected with standard USGS samplers and methods). Furthermore, the variability between either the laser-diffraction or acoustic-backscatter point measurements and the cross-sectionally integrated measurements is typically less than the variability between paired cross-sectionally integrated measurements of concentration and grain size. These observations suggest that more error is introduced during the computation of suspended-sediment loads based on conventional sampling methods than is introduced during the computation of suspended-sediment loads using the laser-acoustic system.