Eocene Adakites In The Cascadia Forearc; Implications For The Position Of The Kula-farallon Ridge

Two recently recognized adakite localities in western Washington provide new constraints on the location of the Kula-Farallon-Cascadia trench-ridge-trench triple junction during the late Eocene. The first locality is in the Bremerton Hills (BH) and consists of dacite dikes that have yielded Ar-Ar ages of 47.8+ or -0.4 Ma and 46.6+ or -1.4 Ma. The second locality is approximately 60 km farther north, near Port Townsend (PT), and consists of proximal block and ash flow deposits that contain dacite clasts. Although not yet dated, the PT deposits are constrained by stratigraphy to be older than mid-to-late Eocene. Rocks in both localities display a wide range of adakite traits (e.g., Al2O3 \u3e 15 wt%, Sr/Y \u3e 30, Yb 100 km west of the estimated position of the trailing edge of the Resurrection Plate at that time, and (2) In both age and location the BH adakites fall between two other rock units that are inferred to mark the KFR-North America intersection: the 51 Ma Walker Creek intrusions on southern Vancouver Island (Breitsprecher et al., 2002) and the approximately 44 Ma Tillamook volcanics in Oregon (Haeussler et al., 2003). The age progression of these three magmatic centers implies southward migration of the KFR relative to North America at a rate of 50-60 km/Ma, perhaps as a consequence of left-slip transform offsets in the KFR and/or a change in travel direction of North America from WSW to more directly W at about 50 Ma (Babcock et al., 1992). Subduction of the KFR may have generated a slab window beneath the present-day Cascades, which would have persisted until sometime after 42 Ma, when the KFR ceased to exist and simple subduction of the Farallon plate was initiated