Strain partitioning in southeastern Alaska: Is the Chatham Strait Fault active?

Abstract

A 1200 km-long transform plate boundary passes through southeastern Alaska and northwestern British Columbia and represents one of the most seismically active, but poorly understood continental margins of North America. Although most of the plate motion is accommodated by the right-lateral Queen Charlotte–Fairweather Fault (QCFF) System, which has produced at least six M > 7 earthquakes since 1920, seismic hazard ass essments also include the Chatham Strait Fault (CSF) as a potentially active, 400 km-long strike slip fault that cuts northward through southeastern Alaska, connecting with the Eastern Denali Fault. Nearly the entire length of the CSF is submerged beneath Chatham Strait and Lynn Canal and has never been systematically imaged using high-resolution marine geophysical approaches. In this study we present an integrated analysis of new marine seismic reflectiondata ac quired across Lynn Canal and tectonic block modeling constrained by data from continuous and campaign GPS sites. Seismic profiles cross the CSF at twelve locations spanning ∼50 km of fault length; they reveal thick (up to 300 m) packages of glaciomarine sedimentary facies emplaced on an unconformity su rface that formed during the Last Glacial Maximum (LGM). Localized warping of post-LGM stratigraphy (∼13.9 kyr B.P. to present) appears to correlate with sediment drape on basement topography and current-controlled deposition. There is no evidence for an active fault along the axis of Lynn Canal in the seismic reflection data. Crustal block models constrained by GPS data allow, but do not require, a maximum slip rate o f 2–3 mm/yr along the CSF; higher slip rates on the CSF result in significant misfit to GPS data in the surrounding region. Based on the combined marine geophysical and GPS observations, it is plausible that the CSF has not generated resolvable coseismic deformation in the last ∼13 ka and that the modern slip-rate is <1 mm/yr. We propose that models for strain transfer between the QCFF and the Denali Fault, and seismic hazard maps in general, may need to be reevaluated.