Interpreting sediment transport data with channel cross section analysis

Suspended sediment load estimation is important for the management of stream environments. However suspended load data are uncommon and scalable models are needed to take maximum advantage of the measurements available. One of the most commonly used models for correlating suspended sediment load is an empirical power law relationship (Qs=aQ^b, Qs: suspended load, Q: flow rate). However, the relationship of log-scaled suspended load to flow rate has multiple exponents for different flow regimes at a given site, so a single power law relationship is not a good fit. Thus we are exploring an alternative approach that employs channel cross section data historically collected by the US Geological Survey during stream gauge calibration. For our research, daily flow and sediment discharge were selected from about 180 possible USGS gauging sites in California. Among those, about 20 sites were relatively unaffected by human activities, and had more than three years of data including near monthly measurements of channel cross section data. From our analysis, a slope break was consistently observed in the relationship of log-scaled suspended load to flow rate as illustrated in Figure 1 for Redwood Creek at Orick, CA. Most of the selected natural sites clearly show this slope break. The slope break corresponds to a transition of flow from a flat, wide stream to flow constrained by steep banks as verified in Figure 2 for the same site. This suggests that physical factors in the streams such as shear stress are affected by this channel morphological change and result in the greater exponent of sediment load during higher flow regime. Figure1. Daily values of measured sediment transport and flow rate reported by USGS between 1970 and 2001.

Figure2. Near monthly values of measured mean water depth and width reported by USGS between 1969 and 1987.