THE BUMP AND GRIND OF CASCADIA FOREARC BLOCKS: EVIDENCE FROM GRAVITY AND MAGNETIC ANOMALIES

Arc-parallel migration of the Cascadia forearc causes relative motion between discrete crustal blocks, and we hypothesize that this motion generates upper-plate earthquakes. The high density and magnetization of mafic rocks (Siletzia) underlying the forearc produce high-amplitude, regional-scale gravity and magnetic anomalies well suited for mapping forearc blocks. We have used simple graphics software to jointly analyze gravity and magnetic data in order to estimate the lateral limits of Siletzia and delineate boundaries between discrete forearc blocks.

With the possible exception of disruption by the Corvallis fault, Siletzia in Oregon appears to behave as a continuous block, extending offshore along the Oregon coast and eastward to underlie the entire Willamette Valley. At the Columbia River, Siletzia is offset along a northwest line that includes the Portland Hills-Clackamas River structural zone. Beneath Washington and Vancouver Island, Siletzia is segmented into eight to ten blocks, variously uplifted or down-dropped, reflecting compression of the Washington forearc ahead of the northward-advancing Oregon block. Block boundaries coincide with major faults, notably the Devils Mountain, Seattle, Tacoma, Olympia, and Doty faults.

Earthquake locations and crustal blocks sometimes coincide in Puget Sound, but the relation remains problematic. For example, nearly all crustal earthquakes in Puget Sound occur northeast of the Olympia fault and east of the Hood Canal fault, but almost no earthquakes are known between the Doty and Olympia faults. Thus, block boundaries do not always control seismicity. The uncertainty between geophysical anomalies and crustal seismicity extends to the region between the Siletzia forearc blocks and the Cascade arc. Based on gravity and magnetic anomalies, we hypothesize that the Doty and Olympia faults are connected to the St. Helens seismic zone, but this is not apparent from the seismicity. The lack of correlation between earthquakes and block boundaries may reflect vastly different time spans: The historic record of earthquakes is brief compared with the geologic record represented in potential-field anomalies. Reconciling the geophysical observations with the earthquake distribution is needed before the block model can be incorporated into regional hazard assessments.