Summary

The analysis of the lower 16 kilometers of the gravel-bed Chetco River and its flood plain focused on understanding bed-material transport and its relation to channel and flood plain morphology. The main study components were (1) detailed mapping and surveying of the valley bottom to document spatial and temporal changes to the channel and flanking bars and flood plains and (2) quantitative investigation of the flux of bed material into and through the study reach. These study components have resulted in a mutually consistent and coherent understanding of the recent history of the active channel and how observed changes may relate to the influx and removal of bed sediment.

Primary Findings

The Chetco River is a wandering gravel-bed river flanked by abundant and large gravel bars formed of coarse bed-material sediment. The upper reaches of the study area are primarily transport zones, with bar positions fixed by valley geometry and the active bars mainly providing transient storage of bed material. The lower river has been aggrading in response to Holocene sea level rise. The Mill Creek and North Fork reaches, between flood-plain kilometers (FPkm) 5 and 10, have historically been the primary area of this aggradation, with consequent active sedimentation and channel migration. Sediment transport capacity is limited in this reach and most net sediment influx into the study area probably accumulates here. A small amount of fine gravel is transported into the Estuary Reach. It is plausible that little gravel-sized bed sediment exits the Chetco River naturally.

The repeat surveys and map analyses indicate an overall reduction in bar area and local decreases in sinuosity, mainly between 1965 and 1995. Some loss of bar area is due to erosion and some has resulted from vegetation colonization and transition to vegetated and developed flood-plain surfaces. Repeat topographic and bathymetric surveys indicate channel incision for large parts of the study reach, with some areas of the North Fork, Mill Creek and Emily reaches incising as much as 2 m. The specific gage analysis at the upstream end of the study reach indicates that recent incision may have followed aggradation culminating in the late 1970s. These observations are consistent with a reduction of sediment supply relative to transport capacity after at least the channel surveys of 1977. Also consistent with this sediment imbalance is the trend of bed coarsening between FPkms 15.3 and 7.7 and the greater degree of armoring for the bars at FPkm 6 and 3 compared to a measurement at the upstream end of the reach.

Multiple and independent analyses, supported by direct measurements of bedload during winter 2008–09, indicate that the mean annual flux of bed material into the study reach is approximately 40,000–100,000 m³/yr after 1970. The year-to-year flux, however, varies tremendously, with some years probably having little or no bed-material entering the study reach, but for some high-flow years, such as 1982 and 1997, as much as 190,000 m³/yr of bed-material enters the reach. For comparison, the estimated annual volume of gravel extracted from the lower Chetco River for commercial aggregate has ranged from 5,000 to 90,000 m³ and averaged about 59,000 m³/yr between 2000 and 2008. Mined volumes, however, probably exceeded 140,000 m³/yr for several years in the late 1970s, greatly surpassing likely replenishment rates.

The historical planform and vertical changes to the lower Chetco River, which almost certainly exist because of a reduced sediment supply relative to transport capacity, have likely resulted from a combination of (1) bed-sediment removal and (2) transient effects as the river has adjusted to the probable large volume of sediment brought in by the flood of 1964. Fully disentangling these factors is not possible with the information available.

Implications Regarding Possible Future Trends and Monitoring Strategies

For a gravel-bed river such as the lower Chetco River, the physical character of the active channel is chiefly the result of bed-material transport processes. At the broad scale, the balance between bed-material transport capacity and sediment supply controls channel morphology. Details of channel conditions depend, however, on various factors including the history of flow and sediment transport, the time lags involved in eroding and depositing sediment, and other local and drainage-basin-scale disturbances that might directly or indirectly affect the channel.

Despite these complexities, if gravel removal exceeds bed-material influx, decreased bar areas and channel incision probably will ensue, similar to the conditions of the late 1970s and 1980s. Such changes likely will be in conjunction with bed coarsening and possibly greater armoring of bar surfaces. Another probable outcome of a sediment deficit would be reduced migration rates, because bar deposition is a major cause of channel migration. Without gravel extraction, aggradation and enhanced channel migration is likely, probably first in the historical sedimentation area of the Mill Creek and North Fork reaches. Because of the low transport capacity in these middle reaches, effects of enhanced sediment supply would probably take longer to affect the Estuary Reach. The time scales of changes depend foremost on sediment influx. A large influx associated with a flood like the one in 1964 could reverse most historical changes during the event. In contrast, the effects of sustained periods of excess transport capacity relative to sediment influx are likely to be manifest over years and decades, and possibly at diminishing rates as the channel and bars coarsen.

Because the sediment balance is a controlling factor, a key aspect of understanding possible effects of various management scenarios on the lower Chetco River is accurate knowledge of the volume of the influx of bed material. For the Chetco River, the bed-material capacity equations applied to the flow record provide seemingly reasonable estimates of bed-material influx to the lower river. This situation offers the opportunity, as long as continuous streamflow measurements are available, to provide annual (or even higher resolution) predictions of the volume of bed-material influx that could be used to guide management actions. Such analyses would be enhanced by a sustained bed-material measurement program, ideally involving at least one or two bedload transport measurements per year, to evaluate the reliability of the transport equations and ultimately develop a site specific bedload transport rating curve.

Another key for improving predictions of channel conditions and documenting effects of management actions is understanding the fate and effects of bed-material sediment entering the reach. Repeat high-resolution topographic and bathymetric surveys of the entire active channel will (1) document the rates at which sediment is moving through the system, (2) allow identification of trends in vertical and planform channel behavior, and (3) provide independent assessment of the sediment influx and transport. Such surveys would ideally be supplemented by periodic bed-material sediment sampling for evaluating bed texture trends. Besides providing for direct and systematic monitoring of the active channel and enhancing understanding of key transport processes, this knowledge may be important for determining relevant management time-scales by providing information on how long it may take the effects of management actions to have desired or detectable outcomes. In contrast, reach scale interrelations between sediment supply and channel and flood-plain characteristics limit the utility of site-specific topographic surveys for predicting and monitoring conditions in a manner responsive to typical management requirements.

From these considerations, an efficient and credible monitoring program would focus mainly on system-wide assessments of sediment influx and channel change. Sediment influx would probably be most reliably evaluated by annual analysis of the streamflow record, ideally supplemented by continued bedload transport measurements in order to improve the accuracy of the influx predictions and to confirm that the capacity-based equations remain appropriate. Continued channel-change assessments could be efficiently based on the LIDAR topographic and estuary and channel surveys from 2008. Repeat high-resolution surveys at 1-year intervals would enable an independent check of the influx estimates as well as allow monitoring of trends in channel and flood-plain conditions. These types of surveys could replace the site specific surveys with little or no loss of information relevant to trend monitoring. Even at lesser intervals, such surveys would probably provide trends and data useful for evaluating planform and vertical changes in the active channel. Monitoring of bed-sediment texture and vegetation could be less frequent (for example, 5–10 years) and would allow evaluation of how these habitat attributes are changing with overall channel condition.

Revised July 2012

First posted May 26, 2010