PALEOCENE - EARLY EOCENE DISPLACEMENT OF TERRANES ALONG THE NORTHERN CORDILLERA MARGIN

The timing and amount of transport of terranes along the northern Cordillera margin remains very controversial, 30+ years after M.C. Blake and Davy Jones championed the concept. Recent revisions in the plate motion models for the northern Paleo-Pacific basin (Doubrovine and Tarduno, 2008) provide a framework to compare latitude vs. time plots for the northern Cordillera between 85 and 40 Ma, when strongly oblique convergence of both the Kula and Farallon plates provided a mechanism for margin-parallel dextral strike-slip. However, in order to use the plate motion models to reconstruct terrane transport accurately one needs to know the location of the triple junction(s) along the margin. We have revisited the paleomagnetic and structural history of the Paleocene Ghost Rocks complex of the Kodiak Islands, AK, to answer the above question. This part of the Chugach accretionary complex contains interbedded pillow basalt and turbidites, which are interpreted to have formed at a ridge-trench-trench triple junction. Our detailed structural studies found that much of the Ghost Rocks that contains volcanic rocks is composed of intact blocks bordered by high-angle fault zones, with steep upright to locally overturned bedding. The majority of the igneous rocks have well-defined magnetizations with unblocking temperatures ranging from 450 to 550° C. The sites from Alitak Bay were corrected for vertical-axis rotations by rotating site-specific bedding strikes to agree with an average regional strike of 250 degrees. Incremental rotations applied to the site-mean directions indicate that best clustering occurs at the optimal rotation; thus the Alitak rocks were magnetized prior to rotation. Volcanics from Kiliuda Bay, using data from Plumley et al 1983 and this study, also have well-defined magnetizations and include sites with opposite polarities which pass the reversal test. The combined rotation- and tilt-corrected data from Alitak with the tilt-corrected data from Kiliuda indicates best clustering occurs at 100% untilting. Thus the data suggest these rocks likely retain their original magnetization. The latitude for where the rocks were incorporated into the prism ranges from ~ 41 - 48° N and supports models for the location of a ridge-trench-trench triple junction along the Cascadia margin in the early Paleocene.