Six years of fluvial response to a large dam removal on the Carmel River, California, USA

Abstract

Measuring river response to dam removal affords a rare, important opportunity to study fluvial response to sediment pulses on a large field scale. We present a before-after/control-impact study of the Carmel River, California, measuring fluvial geomorphic and grain-size evolution over 8 years, six of which postdated removal of a 32 m-high dam (one of the largest dams removed worldwide) and included 11 flow events exceeding the 2-year flood magnitude. We find that the reservoir-sediment pulse following dam removal was relatively small (97 000 ± 24 000 t over 4 years), owing to deliberate reservoir-sediment stabilization. Scaled to the size of the Carmel River watershed and compared against long-term bedrock denudation rates, the post-dam-removal sediment release was slightly less than the annualized long-term sediment export from this basin. New sediment transited >30 km to the river mouth in less than 2 years, assisted by floods 2 and 4 years after dam removal. The sediment pulse fined the downstream riverbed while causing mostly low-magnitude bed-elevation changes: commonly 0.5 to 1 m or smaller, occurring as discontinuous sediment patches or interstitial deposits, aside from the filling and subsequent partial scour of deep pools. There was no major geomorphic reset downstream from the dam site. Geomorphic changes were driven almost entirely by flow rather than by the modest increase in sediment supply, in contrast to recent examples from other large dam removals. The relatively minor disturbance caused by dam removal on the Carmel River is likely analogous to many future dam removals: a relatively small sediment pulse after deliberate limitation of reservoir-sediment erosion, and with an upstream dam remaining in place. Thus, a large dam removal need not lead to major downstream impacts.

Post‐Fire Sediment Yield From a Central California Watershed: Field Measurements and Validation of the WEPP Model

“…The 2017 bathymetric survey showed the new sediment to be 32 cm thick at borehole site B3 on the delta foreset. This newly deposited material was 53% sand with the rest silt and clay, and contained abundant vegetation debris (Figures S2 and S3 in Supporting Information ; East, Tan, et al., 2023). Organic material was abundant generally; the new deposit was 44 cm thick at site B4 and 50 cm thick at site B7, both on the delta foreset, and at those sites the new, uppermost material was also organic‐rich (containing grass, leaves, and rootlets).…”

“…The 1.23‐ml subsamples were treated with bleach solution for 2 hours while being agitated slowly to remove most non‐charcoal organic matter, then wet‐sieved, retaining the fraction coarser than 0.125 mm. This fraction was oven‐dried at 50°C and examined under a binocular microscope at 15× magnification using a gridded background, and the charcoal particles counted (East, Tan, et al., 2023). This method of counting charcoal abundance in a sample of known volume is based on Anderson et al.…”

“…Surveys of the Carmel River 9 km downstream from Los Padres Dam for a separate study (monitoring a control reach above San Clemente reservoir after the 2015 dam removal) found only isolated, small areas of deposition there in 2017 (Harrison et al, 2018). No substantial deposition occurred there over the next two years either (East, Harrison, et al, 2023). We used the sediment composition from the boreholes, discussed above, to define reasonable bounds for sediment density in order to convert volume to mass.…”

“…The abnormally wet conditions of 2017 were caused by an unusual weather pattern later named an atmospheric‐river “family”: repeated storms in close succession comprising highly concentrated vapor transport sustained by tropical heat sources (Fish et al., 2019; Gershunov et al., 2017). In the Carmel River, the extreme rain in winter 2017 led to seven floods (as recorded at USGS gaging station 11143200): two 10‐year flood peaks in January 2017, a ∼40‐year flood (309 m 3 /s) in February 2017, and four other 2‐year flood peaks (U.S. Geological Survey, 2023; see also East, Harrison, et al., 2023). At the gaging station just below Los Padres Dam, the peak flows reached 106 m 3 /s on 8 January 2017, 121 m 3 /s on 10 January, and 116 m 3 /s on 20 February that year (Figure 2).…”

A Watershed Moment for Western U.S. Dams

“…During the removal of 32‐m‐high San Clemente Dam from California's Carmel River in 2015, two‐thirds of the reservoir sediment was sequestered in place to avoid downstream channel aggradation that could have increased flood risk (Harrison et al., 2018). When a 40‐year flood tested that approach the following winter, the modest geomorphic changes downstream of the dam site were attributable just to flooding caused by major rain events, not to effects of sediment from the former reservoir (East et al., 2023). Eight years later the downstream river still has not seen undesired impacts from the dam removal.…”