Search

High-precision GPS geodesy in the Pacific Northwest provides the first synoptic view of the along-strike variation in Cascadia margin kinematics. These results con- strain interfering deformation fields in a region where typical earthquake recurrence intervals are one or more orders of mag- nitude longer than the decades-long history of seismic moni- toring and where geologic studies are sparse. Interseismic strain accumulation contributes greatly to GPS station veloci- ties along the coast. After correction for a simple elastic dis- location model, important residual motions remain, especially south of the international border. The magnitude of northward forearc motion increases southward from western Washington (3-7 mm/yr)to northern and central Oregon (-9 mm/yr), con- sistent with oblique convergence and geologic constraints on permanent deformation. The margin-parallel strain gradient, concentrated in western Washington across the populated Puget Lowlands, compares in magnitude to shortening across the Los Angeles Basin. Thus crustal faulting also contributes to seismic hazard. Farther south in southern Oregon, north- westward velocities reflect the influence of Pacific-North America motion and impingement of the Sierra Nevada block on the Pacific Northwest. In contrast to previous notions, some deformation related to the Eastern California shear zone crosses northernmost California in the vicinity of the Klamath Mountains and feeds out to the Gorda plate margin.

The analysis of 3 years of continuous data from 7 permanent GPS stations along the western Washington section of the Cascadia Subduction Zone indicates that the direction of the observed horizontal velocities (with respect to station DRAO, nominally representing the stable North American continent) is roughly parallel to the relative plate convergence direction of the Juan de Fuca and North America plates and that their magnitude decreases away from the trench. Most of this deformation can be attributed to inter-seismic strain accumulation due to the locking of the thrust interface. When the dislocation model predictions are subtracted from the observed velocities, there is evidence for an additional N-S oriented contraction at a rate of ∼4 mm/yr over a distance of 250 km. This signal presumably represents a more long-term deformation pattern than the periodic accumulation and release of elastic strain connected with subduction earthquakes and is most likely related to the occurrence of shallow earthquakes in western Washington that are characterized by predominantly N-S oriented maximum principal stress.

The Washington Regional Seismograph Network (WRSN) earthquake catalog shows more earthquakes in the day than at night, probably because some local explosions are misidentified as earthquakes. We have implemented a method to help discriminate likely local blasts from local earthquakes. Probable blasts are identified based on time of occurrence (year, month and hour), magnitude, location, and depth. We have defined thirteen areas in Washington and northern Oregon, with enhanced daytime activity probably due to blasts, and have developed specific time, magnitude, and depth criteria for each area to best identify probable blasts. Except in a few regions, most of the smallest, shallowest events occurring during the day in Washington and northern Oregon are probably blasts.

Except for the recent rumblings of a few moderate earthquakes and the eruption of Mt. St. Helen's, all has been relatively quiet on the Pacific Northwestern front. The Cascades region in the Pacific Northwest, a sporadically active earthquake and volcanic zone, still has great seismic potential [Atwater, 1987], as comparisons with other subduction zones around the world have shown [Heaton and Kanamori, 1984]. Recent tsunami propagation models [Satake, 1996] and tree ring studies suggest that the last great Cascadia earthquake occurred in the winter of 1700 A.D. and had a magnitude of −8.9. The North Cascades or Wenatchee earthquake followed in 1872. With an estimated magnitude greater than 7, it was the largest earthquake in the written history of Washington and Oregon.

The intense seismic activity which preceded the 18 May 1980 eruption of Mount St. Helens, Washington, released 2 to 3 × 1018 ergs/day in earthquakes that did not correlate temporally with phreatic eruptions which occurred during the same period. Although the b value and amplitude ratios (long-period/short-period) of the earthquakes vary with time, there are no definitive precursors to the 18 May earthquake and eruption. A Mogi type II frequency-magnitude relation, with critical magnitude Mc = 4.6, constrains the characteristic dimension of the highly stressed region under Mount St. Helens to approximately 3 km, preceding the eruption. A major increase in seismic energy release and a decrease in b value around 1 April 1980 may indicate the first major influx of magma into the upper portion of the volcano.Seismic waves from low-frequency volcanic earthquake have large periods at all epicentral distances. Recordings of volcanic earthquakes from 2 to 4 April 1980 at sites 4 to 9 km from Mount St. Helens show two predominant periods of 0.55 and 1.0 sec. We speculate that seismic signals from the low-frequency volcanic earthquakes have a tectonic origin, but may be modified by pressure oscillations in nearby magma.

Since 1900, more than 290 earthquakes have been reported near Flathead Lake, Montana. Surprisingly, none has exceeded magnitude 5 to 512. Most recent earthquake swarms appear to result from east-west or northwest-southeast extension along short fault segments west and north of the lake. Major normal faults like the Swan and Mission faults east of the lake may pose higher risk, but they appear dormant today. Deformation of sediments in Flathead Lake may be caused by several large earthquakes more than 10,000 years ago but is more probably due to glacial processes accompanying the last retreat of the Cordilleran ice sheet.