USGS Project No. 7360-12400
Alan R. Nelson
Geologic Hazards Team, Central Region
US Geological Survey
MS 966 Box 25046
Denver CO 80225
Collaborators: Harvey Kelsey (Humboldt State University), co-PI on Part A; Eileen Hemphill-Haley (USGS Pacific Marine and Coastal Team), co-PI on Part A; Anne Jennings (University of Colorado), co-PI on Part B. Kelsey and his graduate students are largely funded by the U.S. National Science Foundation.
Element: I and II, PN
Keywords: paleoseismology, recurrence interval, age dating, tsunami
A history of great plate-boundary earthquakes and accompanying tsunamis in the Oregon part of the Cascadia subduction zone is being developed through study of the stratigraphy beneath coastal marshes and lakes. Some buried tidal-marsh soils found in estuarine sequences were submerged due to coseismic subsidence during plate-boundary earthquakes that extended for hundreds of kilometers along the coast. Other soils were submerged as a result of (1) subsidence near shallow faults or folds, either during great earthquakes or during large upper-plate earthquakes, or (2) by nonseismic changes in water levels and sedimentation rates. Histories of plate-boundary earthquakes developed through study of tsunami-deposited sand beds in coastal lakes are being compared with histories from tidal marshes.
The magnitude and recurrence of great plate-boundary earthquakes, and the extent to which they reflect seismic segmentation of the Cascadia subduction zone, remain important issues in seismic hazard assessment in central western North America. But the signs of past coseismic coastal subsidence and tsunamis in tidal-marsh stratigraphic sequences are easily confused with records of local changes in water levels and sedimentation rates, and regional changes in ocean currents. Earthquake magnitude is particularly difficult to determine, partly because conflicting geophysical models of the width of past rupture zones are largely unconstrained by geologic field data. Comparison of tidal-marsh records with sediment records from adjacent low-lying freshwater lakes above the reach of storm surges helps to decipher these complex stratigraphic records, particularly in the southern Cascadia zone where little paleoseismological work has been completed.
Objective A -- To determine the recurrence of plate-boundary earthquakes and accompanying local tsunamis over the past 4000-7000 years in central and southern Oregon through a comparison of stratigraphic records in coastal marshes and lakes.
Objective B - To determine the feasibility of using foraminiferal assemblages in tidal marsh stratigraphic sequences to estimate the width of the rupture zones of past plate-boundary earthquakes and the amount of land-level change during crustal earthquakes beneath Puget Sound.
Objective A -- Work in FY97 focused on completing data analysis for our collaborative study of the 7500-yr record from Bradley Lake in southern Oregon. We completed final sampling of FY94-95 cores from Bradley Lake at NSF core facility in Corvallis and compiled and interpreted the extensive stratigraphic data from cores. We also began to write drafts of two major papers on the Bradley Lake studies. The most recent abstract follows:
The size (magnitude 8 or 9?) and timing of great plate-boundary earthquakes at the Cascadia subduction zone are critical issues in seismic hazard assessments in the central Pacific coastal region of North America. Studies of sudden relative sea-level changes in tidal-wetland stratigraphic sequences have been instrumental in identifying the earthquake hazard, but signs of past coseismic coastal subsidence and tsunamis in such sequences may be confused with records of local changes in water levels and sedimentation rates, and regional changes in ocean currents. Comparison of sedimentary sequences between tidal marshes and adjacent low-lying freshwater lakes above the reach of storm surges helps to decipher these complex stratigraphic records. Lake records reflect more uniform and continuous sedimentation and have stronger contrasts between marine and freshwater fossils than do stratigraphic records from tidal marshes. Correlation of lake records with those of adjacent tidal marshes strengthens a local tsunami interpretation for sandy beds from both types of environments.
A stratigraphic sequence from Bradley Lake, which formed when a dune blocked a stream 28 km north of Cape Blanco in southern Oregon, may contain the longest record of large earthquakes of any site yet discovered at the Cascadia subduction zone. Because the lake (5.5 masl) is about 1.5 m higher than and hundreds of meters landward of historic storm surges, inundation by local tsunamis produced by large, offshore earthquakes are inferred from distinctive beds that indicate as many as 15 times of lake-wide disturbance. In the longest of 14 piston cores and 13 vibracores from the lake, 6 m of largely laminated silty clay and clayey gyttja overlie 0.9 m of peaty sediment; 45 AMS 14C ages on detrital terrestrial macrofossils show that the lake sequence spans about 7200 yr. Distinctive patterns of gray laminae, which record stream floods, allow correlation of cores throughout the lake. Distinctive beds deposited during 12 disturbance events consist of a sand bed and/or coarse organic debris layer overlain by massive brown gyttja, and, in turn, by 1-10 cm of 0.5-1.0-mm-thick light and dark laminae. We infer that sand emplacement was from the west (during at least 6 events), deposition of massive gyttja followed over a period of hours to months, and finely laminated beds accumulated slowly in a meromictic lake for 10-90 yr following each event. The thickness of sand deposited during a disturbance event about 1600 yr ago decreases from 1.7 m at the western end of the lake to 0.1 m, 400 m to the east. Marked increases in the proportion of salt-tolerant diatoms in gyttja beds are further evidence for marine inundation. Missing section and mud clasts in sand beds show erosion of the lake bottom during most events. Assuming fine laminae are annual, intervals between events range from 100-1000 yr. One of the least extensive events occurred 300 yr ago; the most extensive event occurred about 1600 yr ago. Other events date from roughly 900, 1000, 1400, 2700, 3000, 3400, 3700, 4100, 4200, 4500, 4800, 5400, and 6400 yr ago. If most disturbance events record tsunamis, relative sea-level along this part of the Oregon coast must have been close to present for much of the past 7200 years, implying an unusual balance between rates of land uplift and sea-level rise.
Objective B -- New work in FY97 is exploring the feasibility of using fossil intertidal foraminiferal assemblages to measure small (<0.5-1.0 m) differences in the amount of sudden subsidence that occurred during past plate-boundary earthquakes along transects perpendicular to the subduction zone. If differences can be measured with sufficient precision, the width of the rupture zone of past earthquakes may be estimated. Earlier reconnaisance stratigraphic studies have provided little constraint on width or the location of the zero isobase in Oregon. The amount of subsidence near crustal faults in Puget Sound might be estimated in the same way. Several new studies of modern assemblages on Vancouver Island suggest both complications and new opportunities in applying foraminiferal analyses to small-scale sea-level reconstructions. Transects of modern foraminiferal samples were collected in an Alsea Bay, Oregon, marsh by Linda Gerson and she is analyzing them as part of a senior thesis project at the University of Colorado. The modern assemblages will be compared with fossil assemblages from vibracores.
The size and frequency of past great earthquakes along the Cascadia subduction zone remain uncertain, particularly in the southern part of the subduction zone. To decipher the complex paleoseismic record of southern Oregon we are taking a new approach in a cooperative study with colleagues in California (Harvey Kelsey, Humboldt State University) and Oregon (Eileen Hemphill-Haley, PMG)—one that emphasizes field mapping and precise dating of tsunami deposits and accompanying land-level changes in adjacent but contrasting environmental settings. We are studying independent types of evidence of great earthquakes in both brackish-water estuaries and nearby freshwater coastal lakes. Through detailed stratigraphic and laboratory studies we hope to answer questions such as:
Did the tsunami generated by the most recent great earthquake about 300 years ago in southern Washington and northern Oregon also devastate the southern Oregon coast?
How large was this earthquake and what was the inundation level of its attendant tsunami?
How can deposits from local or regional tsunamis be distinguished from the deposits of large storms or floods?
What is the relation of the timing of older earthquakes to those inferred from the tidal-marsh records of northern Oregon, Washington, and California?
Clague, J.J., Naesgaard, E., and Nelson, A.R., 1997, Age and significance of earthquake-induced liquefaction near Vancouver, British Columbia, Canada: Canadian Geotechnical Journal, v. 34, p. 53-62.
McCalpin, J.P., and Nelson, A.R., 1996, Introduction to paleoseismology, in McCalpin, J.P., ed., Paleoseismology: Orlando, Florida, Academic Press, p. 1-32.
Nelson, A.R., Shennan, Ian, and Long, A.J., 1996, Identifying coseismic subsidence in tidal-wetland stratigraphic sequences at the Cascadia subduction zone of western North America: Journal of Geophysical Research, v. 101, B3, p. 6115-6135.
Nelson, A.R., Jennings, A.E., and Kashima, Kaoru, 1996, An earthquake history derived from stratigraphic and microfossil evidence of relative sea-level change at Coos Bay, southern Oregon coast: Geological Society of America Bulletin, v. 108, p. 141-154.
Nelson, A.R., Kelsey, H.M., Hemphill-Haley, Eileen, and Witter, R.C., 1996, A 7500-yr lake record of Cascadia tsunamis in southern coastal Oregon: Geological Society of America Abstracts with Programs, v. 28, no. 5, p.
Nelson, A.R., Ota, Yoko, Umitsu, Masatomo, Kashima, Kaoru, and Matsushima, Yoshiaki, Seismic or hydrodynamic control of rapid late Holocene sea-level rises in southern coastal Oregon, USA?, 1998, The Holocene, (in press).