In the fall of 1805, Lewis and Clark’s Corps of Discovery at last descended on the lower Columbia on the final stretch of its remarkable expedition from the upper Missouri to the Pacific. The men paddled down a broad river studded with islands, laced with backwaters, and choked with shifting, shallow shoals. Night and day from their stockade, Fort Clatsop, at the river’s mouth, the men endured the thunder and roaring of the sea as it crashed over the river bar in an area so treacherous to ships that it was to become known as The Graveyard of the Pacific.

By the early years of the next century, long jetties had been constructed on the shoals and spits at the river’s mouth, narrowing it so that ebb tides would erode a deep entrance channel for safer passage of ships into the Columbia and its teeming ports. In the succeeding 50 years, 11 major dams would be built on the Columbia mainstem, and countless minor dams would span countless tributaries of the nation’s third-largest river. Shoals would be cleared and a shipping channel dredged.

As the millennium approached, alarming events and trends in coastal erosion signaled that the regional dynamics of sediment supply, movement, and deposition might be out of sync in the coastal zone or littoral cell affected by Columbia River sediments. Areas that historically accreted sand were beginning to lose it. Storms breached spits and threatened houses and infrastructure with destruction. Deepening of harbor entrances resulted in concentrated flows, erosion, and subsequent change in the bottom topography of estuary mouths. Shallow-water habitat was lost, and with it, valuable fish and wildlife habitats. Cranberry bogs that formerly were far from the sea had quickly become closer to it. A section of the coast highway had to be abandoned.

Circumstances such as these have important implications for the scale of analysis and the interdisciplinary approach essential for development of lasting solutions to coastal erosion. Quite ironically, they also point to the widespread problems many coastal communities will be facing if global warming trends continue and sea levels rise. Both circumstances ring a wake-up bell calling for scientists, planners, and engineers who make decisions about coastal erosion to consider regional morphology in the management of rivers and coastlines and in the siting and design of sustainable communities at the ocean’s sedimentary edge.

On the southern Washington coast, local municipalities had no background scientific data about coastline conditions to help guide responses to the erosional crises they were experiencing. In fact, historically, much of the coastline in the region had been accreting sand. In one location, accretion had added more than a mile of land to the coastline in historic times. The shift in local economies from resource extraction to tourism had invited coastal development in recent decades. The sudden erosion events caught communities unaware of the scales of coastal change.

Columbia River sediments are delivered to the coastline as a result of complex interrelated dynamics involving the motion of wind, waves, tides, river flow, coastal currents, and the sediments themselves, each influenced by daily, seasonal, decade-long, and longer variations; by shifts in climate; and by ocean floor topography and episodic tectonic events. The inherent complexity of this system and the lack of data made it difficult for local planners, let alone coastal scientists, to peg what was normal, what was changing, and to what degree the changes being experienced should be considered normal.

A regional study grew out of this concern, aimed at illuminating temporal and spatial scales of coastal change in the 99-mi. littoral cell of the Columbia River between Tillamook Head, OR, and Point Grenville, WA. The comprehensive study examines factors in shoreline change variables-climate, sea levels, coastal processes, sediment budget, and human activities-in order to understand and predict coastal change over time.

The results of the study will be used to guide land-use planning and related decisions affecting expenditures of millions of dollars. Spearheaded by the United States Geological Survey and the Washington Department of Ecology and supported by local municipalities, the interdisciplinary study will integrate data to build conceptual models of the sediment budget over multiple scales of time and space. These probabilistic models will yield erosion susceptibility ratings over time, which will be used to identify areas at risk of erosion and flooding (see the Coastal Erosion Study website).

Researchers are looking at geologic and climatic changes that influenced geomorphic processes and materials at a regional scale at the end of the Pleistocene and in the most recent 10,000 years, or the Holocene. Several periods of extensive continental and alpine glaciation took place in the Pacific Northwest during the Pleistocene. Shorelines sank and the sea drowned river mouths, filling them with sediments. Sea levels ebbed as glaciers grew, then land masses bobbed up as they waned, reinvigorating the erosive power of streams. Tectonic plates shifted, affecting shoreline retreat and advance. During dry interglacial periods, wind-blown, glacially pulverized rock was deposited in a thick mantle over vast parts of the region’s landscape. Stream networks incised this material, sluicing it downstream. Ice-age lakes grew to immense proportions where tributaries to the Columbia were dammed by glaciers, then burst their dams in a series of colossal and repeated floods. These vast, regional floods scoured soil down to bedrock, plucked up weathered rock, and swept away the glacial detritus in their paths as they coursed down to the Columbia.

Volcanism also supplied the river with abundant sediments. Two of the five active volcanoes in today’s 260,000-mi.2 Columbia River watershed have sent massive mudflows into the river system since Europeans have been present in the landscape. Geologists now think that Lewis and Clark’s 1805 description of quicksand at the wide and shallow confluence of the Sandy River with the Columbia is an accurate portrait of conditions in the wake of an eruptive episode on Mt. Hood that occurred as recently as 1800. Massive debris and mudflows were generated by the collapse of a growing lava dome and the rapid melting of snow and ice in their paths. These flows mobilized sediments, glacial detritus, and volcaniclastic materials in fast-moving, dense slurries of water and rock that swilled into the Sandy. Today these sediments and others from similar, older events are still mobilized by both rivers. The eruption of Mount St. Helens in May 1980 delivered similar materials to the Columbia via the Toutle River. Researchers for the coastal study have identified signature sands from this event in shoreface materials far north in the Columbia River littoral cell.

Together these events and processes produced a sediment-rich lower river system, a fact reflected in the lower river’s shoals and myriad islands and its extensive estuarine fill. At the river’s mouth, waves and currents move sand in a dominant northward direction, along the shore, where they create the 20-mi. Long Beach Peninsula and beaches north. Seasonal reversals might be responsible for limited sand transport south. A glance at a regional map quickly reveals the long, smooth beaches north of Tillamook Head and those of the Washington coast south of Point Grenville. This is the Columbia River littoral cell, a region whose preanthropogenic sediment budget the researchers estimated to be on the order of 21 million tons a year.

Scientists studied past and present sand transport routes and dynamics and determined that, historically, about two-thirds of the sediment delivered to the Pacific by the Columbia accumulated on the Washington midcontinental shelf. Eleven percent went down submarine canyons to the abyssal plain and was unavailable for transport. Only some 20% was delivered to the inner shelf, beaches, bays, and estuaries. They calculated that anthropogenic effects-the construction of dams and jetties and offshore dredging and upland dumping of dredged Columbia River sand-beginning at about the turn of the last century reduced this sediment load by approximately 70%, leaving just 5 million tons of sand per year or less to be distributed on the shoreface.

But researchers found that far from exhibiting a uniform response to the diminished sand supply, coastal responses were variable in space and time. The north jetty at the Columbia River’s mouth was constructed from 1913 to 1917. At first, it caused rapid erosion of the extensive ebb delta, the treacherous Peacock Spit that had caused so many ships to run aground. The study showed that, by 1926, the jetty had shifted the ebb-tidal delta of the Columbia 3 mi. to the north and 3 mi. offshore. Much of the eroded sand was moved to shore by waves and currents, where rapid accretion of the southern Long Beach Peninsula took place between 1926 and the 1950s. But while the southern portion of the peninsula prograded rapidly, the northern portion eroded during the same period and later began experiencing moderate accretion.

As the shoal shrank, sediment delivery to the Long Beach Peninsula diminished and the rate of progradation slowed. Stratigraphic studies showed that in the late Holocene, the peninsula had served as a sediment transport corridor in quasi-equilibrium, feeding sediment northward in the littoral cell. But with the construction of jetties at the Columbia River and Grays Harbor, it appeared to researchers that the coastlines between the cell’s three major estuaries were receiving less sand began to behave more like closed-system pocket beaches.

Northward, at the entrance to Willapa Bay, the middle of the three estuaries in the littoral cell, the associated change in sand distribution might have triggered the rapid northward migration of the estuary channel. The north bank became known as the most sustained, rapidly eroding stretch of shoreline on the Pacific coast. At the northernmost estuary, Grays Harbor, Ocean Shores had been accreting for 80 years prior to 1997. Yet between 1995 and 1997, the area experienced up to 65 yd. of shoreline recession. The harbor’s ebb-tidal shoal had diminished to the point that researchers questioned whether it was still supplying sand to the adjacent coast.

According to George Kaminsky, manager of the research program for the Washington Department of ecology, the slowdown in shoreline progradation and the onset or acceleration of shoreline recession are likely the results of decreased sediment supply from the ebb-tidal deltas. Further, shoreface steepening near the Grays Harbor south jetty could be diverting longshore-moving sediment to deep water northward and offshore of the Grays Harbor entrance. The distribution of sediment throughout the entire littoral cell has been affected.

Yet many questions remain unanswered, says Kaminsky. Why are there erosion hot spots? Are cross-shore processes becoming more dominant over longshore processes among the four subcells? What is the role of estuarine-fill sediments in the sediment budget of the littoral cell? Are ebb-tidal deltas, the shoreface, and adjacent shorelines approaching equilibrium under the changed conditions? Most important of all: Where will the shoreline be in the future?

Dutch researcher Huib de Vriend of the Delft University of Technology, Netherlands, offers a view on the prediction of large-scale coastal behavior, based on the European project, PACE, in which 60 researchers from Europe, Australia, and the US are participating. de Vriend distinguishes between coastal behavior that is forced and that which is free. Forced behavior is the result of an external input, such as a jetty. Free behavior occurs in specific modes and results from a positive feedback between bed topographical features and the water and sediment movement around them. Examples are small-scale bedforms such as ripples and breaker bar systems.

According to de Vriend, the problem in predicting large-scale coastal change is that computationally demanding research models require input of data derived from small-scale observations such as of individual waves. The amount of data required to drive such models for meaningful scales is so great that the computations are rarely done. But by aggregation, such as the use of a bed roughness indicator instead of a description of the interaction between flow and small-scale bedforms, behaviors can be predicted. He proposes a set of cascading, interlinked spatial and temporal scales that allows behavior-oriented rather than process-oriented modeling. By aggregating compatible scales of time and space, predictions for longer time frames and larger coastal reaches can be made. He cautions that although aggradation of models can overcome predictability limits, the price is a loss of resolution.

Guy Gelfenbaum, study manager for the US Geological Survey, notes that the sediment budget for the Columbia River littoral cell is being evaluated over several time scales, which reflect the seasonal, historical, and geological rates of large-scale coastal evolution. Research on erosion and accretion is done by means of soil profiles, ground-penetrating radar, auger coring, vibra-coring and drilling, and dating of scarps, episodic erosion events, and prehistoric accretion rates. Analysis of shoreline and bathymetric change is yielding regional historical shoreline change rates and accumulation volumes. Beach morphology is monitored by means of cross-shore profiles and sediment samples and analysis of shoreline scarps. Sediment cores, video, and seismic and side-scan sonar are used to analyze inner shelf deposits to establish rates and times of sediment accumulation. And scientists are shining light on the sediment source by compiling data on dredging and disposal, accumulation of sediments in reservoirs, and analysis of midshelf accumulation.

Brian Voigt of the Washington Department of Ecology notes that the results of these studies are being used to develop decision-support tools: maps, a geographic information system database, and reports for use by the coastal management community. A coastal classification scheme will result, whereby realistic scenarios of future coastal conditions can be used to support adaptive management and the development of coastal policy.

“What the study does,” explains Jim Sayce of the City of Long Beach, WA, “is give us the ability to analyze growth. It gives us the ability to make decisions today about when to build and how to build, such that 20 or 30 years from now we’re not looking back and saying, “˜What fools we were!’ That’s the great thing about this study. It is gathering cumulative knowledge that will be used to make our community a really special place-not just in one person’s lifetime, but in many, many people’s lifetimes.”

About the Author

Martha S. Mitchell

Martha S. Mitchell, CPESC, is principal of ClearWater West Inc. (, consultants in erosion and natural resource planning in Portland, OR.