Christopher R. Sherwood

Historical changes in the Columbia River Estuary

Historical changes in the hydrology, sedimentology, and physical oceanography of the Columbia River Estuary have been evaluated with a combination of statistical, cartographic, and numerical-modelling techniques. Comparison of data digitized from US Coast and Geodetic Survey bathymetric surveys conducted in the periods 1867–1875, 1926–1937, and 1949–1958 reveals that large changes in the morphology of the estuary have been caused by navigational improvements (jetties, dredged channels, and pile dikes) and by the diking and filling of much of the wetland area. Lesser changes are attributable to natural shoaling and erosion. There has been roughly a 15% decrease in tidal prism and a net accumulation of about 68 × 106m3 of sediment in the estuary. Large volumes of sediment have been eroded from the entrance region and deposited on the continental shelf and in the balance of the estuary, contributing to formation of new land. The bathymetric data indicate that, ignoring erosion at the entrance, 370 to 485 × 106m3 of sediment has been deposited in the estuary since 1868 at an average rate of about 0.5 cm y−1, roughly 5 times the rate at which sea level has fallen locally since the turn of the century. n nRiverflow data indicate that the seasonal flow cycle of the Columbia River has been significantly altered by regulation and diversion of water for irrigation. The greatest changes have occurred in the last thirty years. Flow variability over periods greater than a month has been significantly damped and the net discharge has been slightly reduced. These changes in riverflow are too recent to be reflected in the available in the available bathymetric data. n nResults from a laterally averaged, multiple-channel, two-dimensional numerical flow model (described in Hamilton, 1990) suggest that the changes in morphology and riverflow have reduced mixing, increased stratification, altered the response to fortnightly (neap-spring) changes in tidal forcing, and decreased the salinity intrusion length and the transport of salt into the estuary. n nThe overall effects of human intervention in the physical processes of the Columbia River Estuary (i.e. decrease in freshwater inflow, tidal prism, and mixing; increase in flushing time and fine sediment deposition, and net accumulation of sediment) are qualitatively similar to those observed in less energetic and more obviously altered estuarine systems. A concurrent reduction in wetland habitats has resulted in an estimated 82% reduction in emergent plant production and a 15% reduction in benthic macroalgae production, a combined production loss of 51,675 metric tons of organic carbon per year. This has been at least partially compensated by a large increase in the supply of riverine detritus derived from freshwater phytoplankton primary production. Comparison of modern and estimated preregulation organic carbon budgets for the estuary indicates a shift from a food web based on comparatively refractory macrodetritus derived from emergent vegetation to one involving more labile microdetritus derived from allochthonous phytoplankton. The shift has been driven by human-induced changes to the physical environment of the estuary. n nWhile this is a relatively comprehensive study of historical physical changes, it is incomplete in that the sediment budget is still uncertain. More precise quantification of the modern estuarine sediment budget will require both a better understanding of the fluvial input and dredging export terms and a sediment tranport model designed to explain historical changes in the sediment budget. Oceanographic studies to better determine the mechanisms leading to the formation of the turbidity maximum are also needed. The combination of cartography and modelling used in this study should be applicable in other systems where large changes in morphology have occurred in historical time.

Sediment-transport events on the northern California continental shelf during the 1990–1991 STRESS experiment

Abstract Measurements of currents and light transmission were made at bottom tripods and moorings arrayed across the northern California continental shelf along the Coastal Ocean Dynamics Experiment (CODE) “C” transect as part of the 1990–1991 Sediment Transport Events on Shelves and Slopes (STRESS) experiment. In combination with meteorological and wave data from the National Data Buoy Center Buoy 46013, these measurements provide information about the physical forcing and resultant resuspension and transport of bottom material between 21 November and 8 March. Sixteen events were identified in the wave, wind and current-meter records for this period. Only two were local storms with southerly winds, but they caused about half of the seasonal net transport. Seven were swell events that combined long-period waves generated by distant storms with local currents. At the 90-m site, swells interacted with the mean northward flow to produce northward transport. During six northerly wind events, upwelling-favorable winds often were sufficient to slow or reverse the mean northward flow and thus caused southward transport. A single current event, which produced moderate southward transport, was observed at the 130-m site. Net transport during the winter experiment was offshore at all sites, northward at the inner- and mid-shelf sites, but southward at the outer-shelf site. The results suggest that local storms with southerly winds may dominate seasonal transport, as on the Washington shelf, but significant transport also can occur during fair weather and during periods of northerly winds.

Determining suspended sediment particle size information from acoustical and optical backscatter measurements

Abstract During the winter of 1990–1991 an Acoustic BackScatter System (ABSS), five Optical Backscatterance Sensors (OBSs) and a Laser In Situ Settling Tube (LISST) were deployed in 90 m of water off the California coast for 3 months as part of the Sediment Transport Events on Shelves and Slopes (STRESS) experiment. By looking at sediment transport events with both optical (OBS) and acoustic (ABSS) sensors, one obtains information about the size of the particles transported as well as their concentration. Specifically, we employ two different methods of estimating “average particle size”. First, we use vertical scattering intensity profile slopes (acoustical and optical) to infer average particle size using a Rouse profile model of the boundary layer and a Stokes law fall velocity assumption. Secondly, we use a combination of optics and acoustics to form a multifrequency (two frequency) inverse for the average particle size. These results are compared to independent observations from the LISST instrument, which measures the particle size spectrum in situ using laser diffraction techniques. Rouse profile based inversions for particle size are found to be in good agreement with the LISST results except during periods of transport event initiation, when the Rouse profile is not expected to be valid. The two frequency inverse, which is boundary layer model independent, worked reasonably during all periods, with average particle sizes correlating well with the LISST estimates. In order to further corroborate the particle size inverses from the acoustical and optical instruments, we also examined size spectra obtained from in situ sediment grab samples and water column samples (suspended sediments), as well as laboratory tank experiments using STRESS sediments. Again, good agreement is noted. The laboratory tank experiment also allowed us to study the acoustical and optical scattering law characteristics of the STRESS sediments. It is seen that, for optics, using the cross sectional area of an equivalent sphere is a very good first approximation whereas for acoustics, which is most sensitive in the region ka ∼ 1, the particle volume itself is best sensed. In concluding, we briefly interpret the history of some STRESS transport events in light of the size distribution and other information available. For one of the events “anomalous” suspended particle size distributions are noted, i.e. larger particles are seen suspended before finer ones. Speculative hypotheses for why this signature is observed are presented.

Columbia river estuary studies: An introduction to the estuary, a brief history, and prior studies

~Fisheries Research Institute, WH-IO, University of Washington, Seattle, WA 98195, USA ZCollege of Oceanography, Oregon State University, Corvallis, OR 97331, USA SDepartment of Botany and Plant Pathology, Oregon State University, Corvallis, OR 97331, USA 4Geophysics Program, AK-50, University of Washington, Seattle, WA 98195, USA 5School of Oceanography, WB-I O, University of Washington, Seattle, WA 98195; current affiliation - Battelle, Pacific Marine Science Laboratories, Sequim, WA 98382, USA

Energetics and sedimentary processes in the Columbia River Estuary

Abstract The Columbia River Estuary is an energetic, sand-bedded system. Mixed tides with a spring tidal range of about 3.6m and a large freshwater discharge produce bottom shear stresses capable of transporting nearly all the sizes of sediment present in the estuary. As a result, the morphology of the estuary is closely related to the fluxes of tidal and fluvial energy. To investigate the relationship between energetics and sedimentary processes, an energy budget based on a one-dimensional harmonic tidal model was developed for the estuary-tidal river system and has been compared with the results of geological studies of the bedforms, large-scale morphology, sediment distribution, and suspended sediment field. The pattern of energy fluxes predicted by the model suggests the division of the system into three regimes: a tidally dominated lower estuary, a mid-estuary energy flux divergence (EFD) minimum region, and a fluvially controlled, tidal-fluvial reach. Model results also show that non-linear interactions between the tidal and fluvial flows are responsible for the suppression of the tides in the tidal-fluvial reach during periods of high flow. Finally, changes in tidal amplitude at the mouth result in a less than proportionate response in tidal elevations at points inside the estuary, because the cubic dependence of dissipation on tidal amplitude damps system responses to such changes. Many features of the observed sediment transport patterns and sedimentary environments can be related to the energy budget. Most of the medium to coarse sands entering the system from the river are permanently retained within the EFD minimum. Much of this deposition takes place upstream of the limits of salinity intrusion and is not, therefore, related to baroclinic circulatory effects. Most of the fine sands and the silts and clays entering the system are not permanently retained. Some of the silts and clays are, however, temporarily retained in a turbidity maximum, whose mean position is near the lower end of the EFD minimum. This position is dictated by the inability of salinity intrusion to extend up the fluvial potential energy gradient. Although some aspects of sedimentary processes cannot readily be related to the energy budget, our model provides a valuable conceptual approach to the dominant long-term sedimentary processes of sand transport and deposition in the Columbia River Estuary that structure the observed sedimentary environments. Biological processes are in turn strongly influenced by these geological patterns.

Sedimentary geology of the Columbia River Estuary

Abstract A multiyear study of the sedimentary geology of the Columbia River Estuary has provided valuable data regarding sediment distribution, bedform distribution, and suspended sediment distribution on spatial and temporal scales that permit delineation of sedimentary environments and insight into the sedimentary processes that have shaped the estuary. In comparison to other more-intensively studied estuaries in North America, the Columbia River estuary has relatively larger tidal range (maximum semidiurnal range of 3.6m) and large riverflow (6,700m3s−1). Variations in riverflow, sediment supply, and tidal flow occur over a range of time scales, making the study of modern processes, as they relate to long-term effects, particularly challenging. Analyses of more than 2000 bottom-sediment grab samples indicate that the bed material of the estuary varies in a relatively narrow range between 0 and 8 phi (1.0 and 0.0039mm) with an overall mean size of 2.5 phi (0.177mm). Sediment size decreases generally in the downstream direction. Sediments from the upriver channels are coarse (1.5–2.0phi; 0.25–0.35mm) and moderately sorted; sediments in the central estuary show wider range and variation in grain size and sorting (1.75–6.0phi; 0.016–0.3mm). Sediment from the entrance region has a mean size of 2.75phi (0.149mm) and is well sorted. Seasonal changes in sediment size distributions occur and are best delineated by those samples containing more than 10% mud (silt plus clay). Sediments containing a significant fine fraction generally occur only in the peripheral bays and in channels isolated from strong currents. Thin deposits of fine sediments are occasionally found in main channels, and the ephemeral nature of these sediments suggest that they may erode and produce the silty rip-up clasts that appear intermittently in the same regions. The distribution of bedforms of various size and shape has been mapped with side-scan sonar during three seasons and at various tidal stages. The presence of bedforms with wavelengths of 6–8m and alternating slip faces about 40cm high indicates that the deeper portion of the entrance region is dominated by tidally reversing lower flow regime sediment transport. Bedforms in the upper reaches of the estuary are much larger, with heights of up to 3m and wavelengths of up to 100m. These bedforms, and the smaller, superimposed bedforms, imply downstream transport under fluvial conditions. In the central estuary, bedforms in the deep portion of the main channels are oriented upriver while those on the shallow flanks of the channels are oriented seaward. The landward limit of upriver bedform transport varies seasonally in response to riverflow fluctuations. A complex array of sedimentary environments exists in the Columbia River estuary. Each environment is influenced by the relative importance of waves, fluvial currents, and tidal currents, as modified by the presence or absence of estuarine circulation, vegetation, or human activity. The importance of these enviroments to the ecosystem of the estuary is discussed in subsequent papers in this volume.

Estimation of stress and bed roughness during storms on the Northern California Shelf

Abstract Turbulent boundary shear stress depends on a roughness length scale which characterizes momentum transfer to the seabed. The effective roughness length is due to physical roughness geometry such as sediment ripples, grain size and processes affecting momentum transfer, such as bedload transport and wave motions. During storms on the California shelf, wave motions dominate the turbulent boundary layer, although feedback through wave-induced bed forms and sediment transport are important. Several intense storms on the Northern California Shelf were monitored with pressure sensors, acoustic current meters and an optical backscatter sensor within 5 m of the bed at 90 m depth. Wave spectra, velocity profiles, turbulent kinetic energy and suspended sediment concentrations were obtained. At 90 m depth, waves were measured in the band of 0.05-0.08 Hz, with nearbed wave velocities of 5–30 cm s −1 . In spite of wave-induced currents of up to three times the mean speed, 30 min average velocities yielded typical logarithmic profiles. Roughness, as indicated by the zero intercept of the logarithmic profiles, z 0 , varied by a factor of 25 throughout the storms, with a maximum of 18 cm, when mean currents were above 5 cm s −1 and the wave amplitude was maximal. Such large increases in z 0 are predicted by models of wave-current interaction. Effects of sediment transport and bed forms are not easily extracted from the data during storms. But, the fairly large non-storm values of ∼0.5cm indicate the effect of bed ripples.

In-Situ Measurements of Particle Settling Velocity on the Northern California Continental-Shelf

Abstract As part of the Sediment TRansport Events on Shelves and Slopes (STRESS) program, a remote optical settling ☐ was deployed on the northern California continental shelf. The device operates by isolating a volume of sediment-laden fluid from the environment and then monitoring its sedimentation behavior with a transmissometer. Results show a bimodal distribution of suspended sediment during low-energy periods on the shelf that reflects the size distribution of bottom sediments. The coarse mode sinks at 0.026 cm s −1 (22 m day −1 ) and the fine mode settles at 0.0025 cm s −1 (2 m day −1 ). Between one-quarter and two-thirds of the total mass resides in the coarse mode. Roughly one-quarter is in the fine mode. The remainder sinks too slowly ( −1 or −1 ) to be resolved during the 18-h measurement cycles. Greatest uncertainty in assigning mass to the various settling velocity classes comes from sensitivity to ill-constrained particle geometry of the conversion from light attenuation to mass. The device failed during higher energy periods, probably due to penetration of fluid into the ☐. Complete isolation of the fluid from the environment would improve the performance of settling ☐es in energetic settings.

Impacts of Watershed Management on Land-Margin Ecosystems: The Columbia River Estuary

Patterns of land use development that have arisen in the Columbia River Basin over the last century are occurring in large river basins worldwide. The consequent modifications of river flow, physical properties, and discharge of sediment and other constituents appear as cumulative effects in land-margin ecosystems, where estuarine processes intercept, entrap, and transform both riverine and oceanic material. These watershed changes alter both the input to the estuary and the fundamental estuarine processes. Our studies of the Columbia River estuary indicate that these human alterations to watersheds can affect the interaction between river flow and the tides, modifying circulation patterns important to estuarine food webs. Mean river flow has decreased approximately 20% since the 19th century; probably 6 to 8% is due to irrigation withdrawal, the remaining 12 to 14% to climate variability. Regulation of river flow has reduced spring freshet flows to about 50% of the natural level, and has increased fall minimum flows by 10 to 50%. The reduction in spring freshets has lowered modern-day sediment input to the estuary to ~25% of that recorded in the latter part of the 19th century. Navigation structures and filling and diking in the lower river and estuary have decreased the tidal prism by about 15%, increased sediment residence time and shoaling, simplified the channel network, and concentrated flow in the navigation channel. In addition to sediment, temperature, organic matter, nutrients, pollutants, and biotic influxes at the estuarine interface, changes in the river discharge regime have modified estuarine stratification, mixing, and residence time. Such modifications have profound effects on sensitive estuarine processes such as those that occur in the estuarine turbidity maximum (ETM), where trapping of suspended material occurs, organic matter is incorporated in a dynamic microbial loop, and important food web linkages to higher level consumers occur. A landscape perspective on the impacts of watershed alterations needs to be included in our emerging regional and global approaches to ecosystem management if land-margin impacts are to be predicted and mediated.

Acoustical and optical backscatter measurements of sediment transport in the 1988–1989 STRESS experiment

Abstract During the 1988–1989 Sediment Transport Events on Shelves and Slopes (STRESS) experiment, a 1-MHz acoustic backscatter system (ABSS), deployed in 90 m of water off the California coast measured vertical profiles of suspended sediment concentration from 1.5 to (nominally) 26 meters above bottom (m.a.b.). An 8-week-long time series was obtained, showing major sediment transport events (storms) in late December and early January. Comparison of the acoustics measurements from 1.5 m.a.b. are made with optical backscatter system (OBS) concentration estimates lower in the boundary layer (0.25 m.a.b.). Correlations between ABSS and OBS concentration measurements and the boundary layer forcing functions (waves, currents, and their non-linear interaction) provided a variety of insights into the nature of the sediment transport of the STRESS site. Transport rates and integrated transport are seen to be dominated by the largest storm events.