[Next Section] [Previous Section] [Contents]

BACKGROUND

Elastic strain

Historically, zones of coastal uplift and subsidence have been observed as an elastic response to great earthquakes within the subduction zones of Chile, Japan, and Alaska (Heaton and Hartzell, 1986). The effect of coseismic subsidence from the great Alaskan earthquake in 1964 (M=8.5) had a profound impact on the susceptibility of flooding in Portage where coseismic subsidence was measured at about 2 m (6.5 ft.) (Plafker and others, 1969). Elastic models for the overriding plate(Vita-Finzi and Mann, 1994) are supported by observations of subsidence and uplift (Plafker and Savage, 1970) in relation to the distance to the subduction zone trench. Briggs (1994) used paleoseismic evidence to hypothesize a zero-isobase crossing the southern Oregon coast. Landward of the first (trenchward) zero-isobase the coseismic response should be subsidence, whereas seaward of the first zero-isobase the coseismic response should be uplift (Figure 2). A second zero-isobase located landward of the first might have minor deformation, below the resolution of marsh studies (see below). Accumulating strain (interseismic) is currently occurring in the CSZ measured as the shortening of the overriding plate (Dragert and others, 1994) and as uplift of the coastal regions of California, Oregon, Washington (Kelsey and others, 1994; Mitchell and others, 1994).

Paleoseismic evidence

Evidence of coseismic subsidence along the CSZ was first documented by Atwater (1987) in southwestern Washington. The coseismic subsidence interpretations are based on wetland cutbank outcrops where a buried organic-rich "peaty" soil is overlain by a barren mud or rooted mud. Buried peaty deposits of marsh and forest soils record a sea level change as small as 0.5 m to perhaps greater than 2.0 meters (Peterson and others, 1996). Buried marsh deposits in Oregon, northern California, Washington, and Vancouver, B.C. suggest that the effects of coseismic submergence occur on regional and local scales(Darienzo and Peterson, 1990; Clague and Bobrowsky, 1994; Jacoby and others, 1995). Correlation of the observed buried wetland stratigraphy continues to be constrained (Darienzo and Peterson, 1995; Atwater and others, 1995). Correlation of deposits relies on radiocarbon age determinations of buried peats and submergence-associated detritus (Nelson and others, 1995). However, precision tree-ring dating is utilized at localities where submergence was great enough to kill Red Cedar or Sitka Spruce indicating that the trees died within a few years after the initial inundation (Atwater and Yamaguchi, 1991; Jacoby and others, 1995). Most recently reported, historical far-field tsunami records in Japan suggest the date of the last Cascadia earthquake occurred at 1700 AD (Satake and others, 1996).

Criteria for estimating the amount of coseismic subsidence

The criteria that are used to estimate the amount of coseismic subsidence in buried wetland deposits include: (1) plant macrofossils, (2) amount of organic content, (3) abruptness of the buried wetland contact, and (4) microfossils. These criteria have been used in various combinations to estimate paleosubsidence.

Communities of modern wetland plants of known tidal elevations are used as analogues to interpret paleotidal elevations from plant macrofossils found in buried deposits (Atwater, 1992). Plant rhizomes that are generally diagnostic of specific tidal elevations include; Triglochin maritima, Salicornia virginica (low marsh); Grindelia stricta, Carex lyngbyei, Deschampsia caespitosa, Potentilla pacifica (high marsh), and Picea sitchensis (forest).

Another technique of paleosubsidence estimation is based on the change in organic content (Darienzo, 1991). Specifically, the amount of peat is assumed to be proportional to wetland development above the reach of moderate tides (Redfield, 1972). Peat to siliciclastic sediment ratios increase from colonizing to high marsh settings (Figure 3). Organic content estimates are completed in two ways: (1) in the field using a visual estimate, and (2) in the laboratory using loss on ignition (see methods) (Franklin and others, 1973).

Some debate exists regarding the interpretation of the abruptness or thickness of the contact of mud over peat. Burial contact thickness is split into intervals of ~5 mm, ~1mm, and <1 mm to interpret the rapidity of paleosubsidence (Shennan and others, 1995; Nelson and others, 1995). The interpretation may not account for abrupt burial where some plants within a wetland are not immediately killed from subsidence. For example, a forest environment that abruptly subsides to a low marsh environment might take several years for complete recolonization by low marsh plants (Peterson and others, 1996). Lithologically, this example might be recognized as a change from a muddy peat to an overlying rooted mud, with a contact perhaps greater than 1 cm thick.

Oxidation of emergent peaty deposits often results in faint soils, making the identification of buried wetland horizons difficult when based solely on macrofossils. These faint soils may be devoid of the abundant plant remains that they contained prior to oxidation during wetland emergence. To support the identification of supratidal settings in such deposits, diatoms are used to establish fresh water conditions because many diatom species consist of robust silica valves that are resistant to oxidation (Darienzo, 1991; Briggs, 1994; Hemphill-Haley, 1995b; Shennan and others; 1995). Buried microfossil assemblages are compared with studies of modern wetland microfossil zonation (Riznyk, 1973; Rao and Lewin, 1976; Foged, 1978; Whiting and McIntire, 1985) to determine the amount of past subsidence. Recent sampling of modern marsh environments is performed proximal to buried deposits so that environmental differences between the modern and buried settings are minimized (Hemphill-Haley, 1993; Shennan and others; 1995). Ranges for past subsidence based on diatom evidence are from 0.8 to 3.0 m (2.6 to 9.7 ft.) in Willapa Bay (Hemphill-Haley, 1995a).

It is important to recognize sites that reflect subsidence due to local settlement rather than tectonic subsidence. In many of the estuaries, buried wetland deposit localities are underlain by many meters of unconsolidated strata. At these sites it may be difficult to differentiate between the tectonic component of subsidence and that of settlement, however minimal settlement may be (National Academy of Sciences, 1968). For this reason, it is important to locate those areas where potential settlement can be constrained. Such areas include areas of shallow marsh development over existing bedrock or consolidated Pleistocene deposits.

The study area consists of four estuaries chosen to test paleotide-level criteria, and two estuaries chosen to evaluate the flooding potential due to coseismic subsidence (Figure 4). The two estuaries selected for the flooding evaluation, Grays Harbor and the Necanicum Estuary, differ greatly in size and hydrography. In comparison, historic flooding in Tillamook Bay and Siletz Bay are more directly affected by river discharge. For example, observations of flooding at riverine flood plains in these bays in 1996 showed little or no discernible change between tides (Tom Ascher, pers. communication). The four estuaries taken together provide a range of hydrologic and geologic conditions necessary to evaluate the regional application of paleotide-level indicators (see above).

Estuary descriptions presented below include general location by proximity to the Columbia River mouth at the Washington-Oregon state border. The tributaries are noted as well as the range of mean and diurnal tides near the wetland survey and deposit sample sites. The mean tide level range is calculated by subtracting mean low water (MLW) from mean high water (MHW) (National Oceanic and Atmospheric Administration, 1996). In contrast, the diurnal tide range is calculated by subtracting mean lower low water (MLLW) from mean higher high water (MHHW). The diurnal tide range is greater than the mean range and reflects more extreme conditions. In addition to tidal ranges, a brief description of the abundance and quality of paleosubsidence evidence is given for each estuary.

The Grays Harbor estuary (Figure 5) is located approximately 96 km (58 mi.) north of the Columbia River. It has 5 major tributaries, the Chehalis River, Johns River, Humptulips River, Hoquiam River, and the Wishkah River. Tides measured at the city of Aberdeen have mean and diurnal tide ranges of 2.4 m (7.8 ft) and 3.1 m (10.0 ft), respectively (National Oceanic Atmospheric Agency, 1996). Records of recent storm surge along the Washington coast show an increase of 1.2 m (3.9 ft.) to 1.5 m (4.9 ft.) above predicted tides (Phipps, 1990).

Buried wetland deposits for the most recent burial are often observed in cutbanks (Atwater, 1988a). These cutbank exposures have been traced laterally on the order of one kilometer in some localities of Grays Harbor (Atwater, 1992).

The Necanicum Estuary (Figure 6) is located approximately 42 km (25 mi.) south of the Columbia River. It is fed by two major tributaries, the Necanicum River and Neawanna Creek. The Necanicum River is 34 km (20 mi.) long and runs nearly parallel to Neawanna Creek, which is due east of the Necanicum River. The is only 11 km (7 mi.) long. Heads of tide for the Necanicum River and the Neawanna Creek are located at river 4.3 km (2.6 mi.) and 6.1 km (3.7 mi.), respectively (Darienzo, 1991). The Necanicum Estuary experiences mean and diurnal tidal ranges of 1.4 m (4.7 ft.) and 1.8 m (5.8 ft.), respectively at the 12th Avenue Bridge in Seaside (National Oceanic and Atmospheric Administration, 1996).

Tillamook Bay (Figure 7) is approximately 91 km (55 mi.) south of the Columbia River. Five tributaries drain into the bay: Miami River, Kilchis River, Wilson River, Trask River, and Tillamook River (from north to south). The mean and diurnal tidal ranges in Bay City are 1.7 m (5.4 ft.) and 2.2 m (7.1 ft), respectively (National Oceanic Atmospheric Agency, 1996). Further south at the Hoquarten Slough within the town of Tillamook, mean and diurnal tides are 1.6 m (5.2 ft.) and 2.0 m (6.6 ft.), respectively. Exposures of the most recently buried marsh are rarely visible in cutbanks, especially in the southern portion of the bay. High rates of sedimentation in the southeast portion of the bay have resulted in rapid progradation of a bay head delta. The progradation has made it difficult to locate the buried wetland records in the southern portion of Tillamook Bay. Cutbank exposures are rare in the bay due to diking.

Siletz Bay (Figure 8) is approximately 192 km (115 mi.) south of the Columbia River and is fed by the Siletz River, Drift Creek, and Schooner Creek. Mean and diurnal tides in Taft are 1.5 m (5.0 ft.) and 2.0 m (6.6 ft.), respectively (National Oceanic and Atmospheric Administration, 1996). Cutbank exposures of buried wetlands are observed only along Millport Slough (Darienzo, 1991). Much of the stratigraphy for buried deposits in Siletz Bay is constrained only through coring. Extensive diking has reduced modern marsh exposures to localities proximal to tributary mouths.

Flood insurance studies have been conducted for three of the estuaries (Grays Harbor, Necanicum Estuary, and Tillamook Bay) in this study (Army Corp of Engineers, 1971; Soil Conservation Service, 1979; Flood Emergency Management Agency, 1990;). Each study is based on an assumed constant mean sea level for the 10 and 100 year flood. Prediction of flood elevations for each estuary is dependent upon hydrographic conditions. The flooding hazard in Grays Harbor is dominated by storm surge and tidal influences (Army Corps of Engineers, 1971), whereas the flooding hazards for the Tillamook and Siletz Bay areas are mainly controlled by river discharge.

An evaluation of potential flooding hazard has been determined by the Army Corps of Engineers for the Grays Harbor area, specifically the populous areas including Aberdeen, Hoquiam, and Cosmopolis, (1971). Flood hazard maps constructed use an "intermediate regional flood" (100 yr. flood) that consists of discharges from the Chehalis, Wishkah, and Hoquiam Rivers, coupled with assumed maximum tidal conditions. The predicted flood height is 3.0 m (10.0 ft., MSL) near Aberdeen, Washington. A flood of this magnitude is recorded occurring in 1933 (Army Corp of Engineers, 1971). The intermediate regional flood is defined as having a frequency of occurrence of once in 100 years. The discharge estimates are complicated by a lack of nearby stream gauge data for any of the rivers. The nearest gauge on the Chehalis River is located in the city of Porter and has been recording since 1952. However, at Porter the river is relatively unaffected by tides and may not record flooding from storm surge events.

Figure 8. Map of Schooner Creek (SC) and Siletz River (SR) sites in Siletz Bay, Oregon. Asterisk (*) and cross symbols (+) differentiate between modern marsh survey and buried wetland sites, respectively. UTM coordinates are in zone 10.

A flood insurance study for the city of Seaside/Necanicum estuary area (Soil Conservation Service, 1979) utilized a two-part hydraulic analysis i.e., coastal and riverine. A stillwater level is calculated, combining the astronomical tide height and storm surge height. Storm surge was calculated using the COAST program (CH₂M Hill, 1975). An offshore water depth at each point in a two-dimensional grid is entered into the program with one side of the grid representing the coastline. Atmospheric and gradient pressure fields from representative surge producing storms for the last 32 years are placed into the COAST model to calculate storm surge magnitudes. The study’s calculations represent storm surge magnitudes for northwest Oregon.

Due to the lack of recorded data for river discharge in the Necanicum Estuary, a regional analysis of 70 stream gages in the northwestern US coastal area was performed by the Soil Conservation Service (1979). The components of river discharge and tidal influence are used to create a joint frequency probability curve to predict the 10 and 100-year flood heights. The predicted flood height of the Necanicum mouth and head of tide at Neawanna Creek are 3.29 m (10.8 ft.) and 3.81 m (12.5) above MSL, respectively (Soil Conservation service, 1976, 1979). Observed flooding in the Necanicum Estuary has reached 10 yr. flooding heights from storm surge events backing up the river in 1932, 1967, and 1982 (Tom Horning, pers. communication).

[Next Section] [Previous Section] [Contents]