A previous report by the authors described sediment sampling and drilling by the U.S. Geological Survey (USGS) beside the American and Sacramento Rivers near Sacramento, California, in support of a U.S. Army Corps of Engineers project focused on regional flood control. The drilling was performed to define lithology, extract samples for laboratory testing, and perform borehole erosion tests (BETs). The U.S. Department of Agriculture (USDA) performed jet erodibility tests (JETs) near each drilling site, and a team from Texas A&M University performed laboratory tests with an erosion function apparatus (EFA). Collectively, the effort was intended to reveal spatial variations in sediment erodibility and provide data for use in a model to simulate morphological response to a major flood. The data collected by the USGS are available in a public data release.

This report, developed in cooperation with the U.S. Army Corps of Engineers, provides comparisons of the three types of measurements of the erodibility of riverbed sediments. The BET is performed in the field and reveals erodibility of sediments below the bed surface. The JET is likewise performed in the field but reveals only erodibility of exposed sediments. The EFA test is done in the laboratory and was performed on soils extracted from different depths beneath the bed surface, in many cases reconstituted for laboratory testing. Tests were performed at nominally similar locations but differed by meters to tens of meters in horizontal locations.

The comparison was undertaken to investigate differences among results obtained by the individual measurement approaches and to elucidate pros and cons of each method. The critical shear stress to initiate erosion and the rate of change of erosion rate per unit increase of excess shear stress, sometimes referred to as the erosion coefficient, served as the primary basis for comparison. The three test methods in some cases resulted in order of magnitude differences in estimates of these parameters. Some differences could be attributed to variances in site location or result from testing surface sediment versus a deeper layer, but systematic differences are also evident in the results. The tests performed in the laboratory using the EFA resulted in much lower values of critical shear stress and much higher values of the erosion coefficient compared to the JET tests performed by the USDA team on surface sediments. Critical shear stress was poorly resolved in the BET results because of the limited number of results per site, but the erosion coefficients derived from BET results were systematically lower than those obtained using the EFA.

A new, simplified approach is also proposed to estimate the increase in channel cross-sectional area during a large flood, given data describing the initial river cross section, riverbed erodibility parameters, and peak flood discharge and duration. The model runs until the cross section erodes to an equilibrium condition or the flood ends. Output describes the area of the cross section at the end of the simulation and the time required to reach equilibrium if it was reached within the simulated period. The model assumes unique, constant values for both the critical shear stress and the erosion coefficient and represents the fluid mechanics in a simplified way, making it of limited value for quantitative predictions. It does, however, provide an indication of which cross sections are most likely to undergo the greatest change in the design event and can be used to investigate sensitivity of erosion predictions to variability in sediment erodibility measurements.