ALBERT

Description / Table of Contents: PREFACE The aim of this volume is two-fold. At the more pragmatic level, it is to help answer the many questions about the structure of the Pacific continental margin of North America, which have arisen over the years as a result of continuing field mapping and geophysical surveys. The second objective is methodological - to illustrate the irreplaceable role of geological information among the various data sets used in earth-science studies. The need to address these issues became apparent to the author during the several years he spent taking part in geological and geophysical studies on the west coast of Canada. All too often, results of geologic field mapping disagreed with tectonic predictions from too-straightforward local applications of global plate reconstructions, which due to their generality do not always take a full account of specific character of particular regions. To be sure, the global approach has during the last q~/artercentury greatly expanded the vision of geoscientists, previously restricted to continental regions. However, a negative by-product of this expansion has been a decline of attention paid to local information, as tectonic studies have increasingly relied on simply fitting the development of a particular region into this or that prefabricated tectonic template. Direct geological observations have limitations of their own. The observer in most cases deals with products of geologic processes, rather than with the processes themselves. Field mapping provides local information, and many years of effort are needed before a regional overview becomes possible. Geologic mapping is restricted to the ground surface, and even the deepest drillholes cannot sample more than the outermost shell of the Earth. The factual side of geologic mapping is usually limited to determination of rock types and their relationships in areas of exposure. Conclusions about the three-dimensional structure of a region and its evolution are still mostly inferential. Broad incorporation into geological studies of geophysical data, assisted by ever-more-sophisticated modern computers, provides a huge volume of information unobtainable in other ways. Geophysical methods quickly afford regional coverage or images of the Earth's deep interior. Geophysical methods have prompted the application in geological sciences of methodologies borrowed from exact sciences, such as mathematics and physics. Particularly important has been quantitative modeling, which allows a scientist to use the known parameters of a system to predict others. But in taking this approach too far, one encounters a dangerous pitfall. A model is a simplified representation of a natural phenomenon. The quality of this or that representation is relative, and a representation is never perfect. To incorporate all characteristics of a geologic phenomenon, in a parametrized form, into a numerical or physical imitation is impossible. This requires one to rely on simplifying assumptions, and a model is no better than the assumptions at its base. Unrealistic assumptions lead to unrealistic models. When a disagreement arises between model predictions and observations - such as those from geologic field mapping - a modeler may be tempted to downplay the differences or the significance of the offending observations. It becomes tempting to underestimate the role of an experienced geologist as a principal arbiter of the realism of a model. But it is geological data and geological control that provide the ultimate means of testing abstract models. From this methodological position, the present study of the western North American continental margin is organized as follows: 1. Geological information, available from field mapping and drilling, is gathered and summarized. 2. Current geophysical models for this region are considered, with particular attention to their underlying assumptions. 3. The available data, geological and geophysical, are synthesized into an internally consistent geologic-evolution concept. 4. This concept is tested by comparison with direct geological observations from field mapping and drilling. Because most current data sets and models cover northwestern Washington and western British Columbia, particular attention was paid to these areas. Fortunately, these areas contain many keys that help understand the structure of the entire western North American continental margin, which has baffled scientists for decades. The author does not claim to have resolved all these problems, but he does believe he has made a useful contribution to understanding continental-oceanic plate interrelations at this continental margin. Rigidity of lithospheric plates is a critical assumption in current models of plate evolution. The lithophere of a plate is created at spreading centers manifested in the global system of mid-ocean ridges. It moves away from the place of its birth towards boundaries with other plates, with which it can interact in a variety of ways. Some interactions are of strike-slip type, with two plates simply sliding past each other. However, to compensate for the creation of new lithosphere at spreading centers, older lithosphere at some plate boundaries descends into the mantle as it is overriden by other plates. At such plate boundaries lie subduction zones. If both regimes occur along a single plate boundary, the transition between them must be abrupt. Unless it can be tied to a change in orientation of the boundary, it must be associated with a junction of not two, but three different plates. Such a template was used to interpret the structure and tectonic evolution of the western North American continental margin in the late 1960s and thereafter (Atwater, 1970; McManus et al., 1972; Barr and Chase, 1974; Riddihough and Hyndman, 1976). To satisfy the principles of rigid-plate tectonics, both regimes have to exist along this continental margin. Also needed in rigid-plate reconstructions is a plate triple junction somewhere between the areas of proven ongoing subduction (in Oregon and southern Washington) and transform plate motion (along the southeastern Alaska margin; Atwater, 1970; McManus et al., 1972). Such a triple junction has been placed off Queen Charlotte Sound offshore British Columbia (Keen and Hyndman, 1979; Riddihough et al., 1983), where a spreading center has been postulated between the Pacific and Explorer oceanic plates (Hyndman et al. 1979; Riddihough, 1984). Off northern Vancouver Island, a transform boundary between the Explorer and Juan de Fuca oceanic plates has been postulated, but both these plates are assumed to be subducting beneath Vancouver Island (Hyndman et al., 1979; Riddihough and Hyndman, 1989)o With the assumed universality of the rigid-plate model, "broad similarity" has been suggested between the geology of western Oregon and that of western British Columbia, and the Cascadia zone of active subduction has been extended as far north as the mouth of Queen Charlotte Sound (Riddihough, 1979, 1984). An accretionary sedimentary prism (Yorath, 1980) - or even an accretionary complex containing several exotic "terranes" (Davis and Hyndman, 1989) - has been postulated off Vancouver Island. Geological observations onshore and offshore (Shouldice, 1971; Tiffin et al., 1972) have come to be considered too "surficial" to be of major consequence for large-scale tectonic modeling (Yorath et al., 1985a,b; Yorath, 1987). Variants of the principal geophysical model for this area during the last decade (Clowes et al., 1987; Hyndman et alo, 1990; Spence et al. 1991; Yuan et al., 1992; Dehler and Clowes, 1992) have become increasingly distant from geological observations. As new model variants emerged, they were checked for internal consistency, compatibility with neighboring local models and fidelity to the overall assumed tectonic picture. However, detailed geological work continued, and many of its results proved incompatible with the conventional wisdom (Gehrels, 1990; Babcock et al., 1992, 1994; Allan et al., 1993; Lyatsky, 1993a). Importantly, questions arose about the applicability in this region of the conventional, simple rigid-plate assumption, as it was shown to be unable to account for all the geological and geophysical peculiarities in some areas (Carbotte et al., 1989; Allan et al., 1993; Davis and Currie, 1993). New solutions were made necessary by new findings and by rediscovery of forgotten old data (see Lyatsky et al., 1991; Lyatsky, 1993b). Without aiming to resolve all the outstanding debates, tectonic implications of the geologic mapping and drilling results in this region are considered in the following chapters. These results are integrated with geochemical and geophysical data. Interpretations of these data, made by this author and by other workers, are verified by geological observations and by geologically plausible extrapolations from these observations. In searching for solutions consistent with all the information, the author has restricted himself to analyzing continental-crust structures along this continental margin. He believes, however, that future models for the offshore regions of the northeastern Pacific should consider the results obtained herein.