Calling the Cedar River’s Bluff: A study of bed elevation changes and characterizing bluff grain size distributions in the Cedar River upstream of a depositional zone in the City of Renton, Washington

Abstract

The Cedar River, a 477 km2 basin draining the Western Cascade Range, has experienced meters of channel aggradation over the last century in the lower three river kilometers (rkm), which has reduced channel conveyance and increased flooding hazards in Renton, Washington. To understand potential drivers of the channel aggradation in Renton this study investigates potential sediment supply from upstream bed incision over the last 18 years and the proportion of sediment delivered from a channel-marginal bluff that has the potential to contribute to the aggrading bed in Renton. In 1916 the Masonry Dam was constructed, which reduced flows and impounded sediment upstream, possibly causing downstream channel incision. In the 1960’s, levees and revetments were constructed along 60% of the lower 36 kilometers of the Cedar River banks. A gravel supply study in 2002 suggests that channel incision occurred following flow reduction from the dam and the construction of the levees and assisted in preventing the river from flooding the width of the valley floodplain (Perkins et al., 2002). Previous studies have assessed bed elevation changes in the first 35 rkm of the Cedar River between 2000 and 2012. The first objective of this study analyzes bed elevation over the last 18 years to identify whether channel incision is a potential sediment source for the aggrading bed in Renton. Additionally, it analyses cross sections within recent levee setback projects and a landslide zone that have changed the shape of the channel. By comparing existing cross-sectional data from 2000 and 2003 to 2012 and 2012 to 2018, this study reveals there has been no net change in bed elevation. Therefore, bed incision has not contributed to the aggradation in Renton over the last 18 years. During this time, a landslide in 2001 and two levee setback projects altered the active channel and opened the river back to its floodplain. Although these alterations to the channel have higher variability in bed elevation changes, no net average elevation changes were detected. The 2001 landslide between rkm 7.7 and 8.3 has an average cross section change in bed elevation of 0.02 m/yr with a standard error of (±) 0.00 m/yr from 2000 to 2012 and -0.04 ± 0.04 m/yr from 2012 to 2018 (number of cross sections (n) =6). The Cedar River Rapids levee setback between rkm 11.5 and 11.9 has an average bed elevation change of -0.01 ± 0.02 m/yr from 2000 to 2012 right after it was constructed in 2012 and -0.00 ± 0.03 m/yr from 2012 to 2018 (n=4). Rainbow Bend, the second levee setback zone upstream, constructed in 2014 at rkm 17.2 to 18.5, has an average elevation change of 0.02 ± 0.04 m/yr between 2012 and 2018 (n=9). All sub-reaches are not statistically different from a zero net elevation change using a two-tailed t-test at a confidence level of 95%. The second objective of this study is to estimate the grain size distribution of a bluff previously estimated to have the largest sediment supply to the lower Cedar River and evaluate the proportion of grain sizes in the eroding bluff contributing to channel aggradation in Renton. Perkins et al. (2002) reported that 49% of the sediment supplied to the lower Cedar River is from bluff retreat and visually estimated the proportion of gravel in each contributing bluff. This study tests a grain size image detection method to obtain a more accurate estimate of the bluff grain size distribution than the previous visual estimates. Based on the resolution of the images collected in the field, the optical granulometry methodology used in this study was determined to confidently detect particle sizes larger than approximately 95 mm and therefore failed to give a more accurate measurement of the smaller gravel distributions in the bluff. Smaller grain sizes were measured using grab sample sieve analysis and facies mapping of sand lenses in the bluff, estimating 21% sand in the bluff. The sand lenses mapped in the images gave a relatively objective measurement of the minimum proportion of sand in the bluff from sand lens deposits than visual estimates. The grab sample grain sizes, although not representative of the entire bluff, reveal a total sand proportion 16% higher proportion of sand than previous estimated. The bluff estimates are compared to existing bed material samples in Renton. The estimated bluff gravel proportion, after assuming a fourth of the material is lost to attrition, contributes approximately 60% to the gravel to the upper 2.5 to 2 rkm in Renton. The sand proportions of the bluff contribute to the lower reaches in Renton. Therefore, the bluff is an important source of sediment depositing in Renton and to the associated increase in flood frequency.