ALBERT

Description: A model for interstitial silica concentrations is derived, incorporating biological mixing of sediments. This model predicts concentrations and gradients and can account for the observed geographical variations in interstitial silica on the basis of a dynamic balance between solution of silica particles and diffusion from the sediments. The flux of particulate biogenous silica into the sediments is confirmed as an important parameter controlling interstitial silica concentrations. Biological mixing of sea floor sediments also has an important influence on interstitial composition by modifyirig the depth at which dissolving particles react. Faster mixing raises the interstitial concentration. The rate at which siliceous particles dissolve also plays a role; the slower they dissolve, the greater the interstitial silica concentration. Measurements on near‐bottom waters of the Atlantic show no consistent gradients in dissolved silica, but antarctic bottom water seems significantly more variable in the benthic boundary layer than in the water mass above or in the benthic zone of North Atlantic deep water.

Description: Biological mixing in deep‐sea sediments is described in terms of a time‐dependent eddy diffusion model where mixing takes place to a depth L at constant eddy diffusivity D. The differential equation that describes this model has been solved for an impulse source of tracer delivered to the plane surface that forms the top of the mixed layer. The solution then serves as a Green's function, which can be used to determine the distribution of tracer in depth and in time for a surface input of tracer specified as any arbitrary function of time. The characteristic properties of the solution are dependent on the dimensionless parameter D/Lυ, where υ is the sedimentation rate. If D/Lυ is greater than 10, the surface layer becomes homogeneous, and the model is identical to the homogeneous layer model proposed by Berger and Heath (1968). If D/Lυ is less than 0.1, little mixing can take place before the sediments are buried, and so the surface concentration propagates downward into the sediments with little dispersion. For all values of D/Lυ the weighted mean depth of the concentration distribution is the depth at which an impulse source would be found in the sediment if no mixing had taken place. The microtektite data of Glass (1969, 1972) and Glass et al. (1973) indicate that abyssal sediments are mixed from the surface to a maximum mixing depth that ranges between 17 and 40 cm below the surface. Mixing occurs at rates between 1 and 100 cm2 kyr−1. Higher mixing rates may occur nearer the surface, but microtektite distributions cannot be used to estimate these rates in the presence of the deeper, slower mixing. Estimates for D based on dimensional analysis of sediment reworking rates for nearshore organisms (103–106 cm2 kyr−1) are used to predict abyssal mixing rates between 1 and 103 cm2 kyr−1 by invoking the assumption that mixing is proportional to biomass. Plutonium distributions in deep‐sea sediments (Noshkin and Bowen, 1973) indicate abyssal mixing rates ranging from 100 to 400 cm2 kyr−1.

Description: The eastern boundary of the Caroline plate, in the western equatorial Pacific, is composed of three structural provinces distinguished primarily on the basis of morphology. Each province shows evidence for convergence between the Caroline and Pacific plates though the structural style varies considerably between each province. Most notably, the sense of underthrusting appears to change along the boundary at about 3°N. To the south, at the Mussau System, Caroline lithosphere underthrusts beneath the Mussau Ridge (which is part of the Pacific plate), while to the north the Caroline plate appears to overthrust the Pacific plate. Recently collected seismic reflection profiles across each province documents the structural changes along and across strike of the Caroline-Pacific plate boundary. With this information, we estimate that a minimum of approximately 4 km of crustal shortening has occurred at about 5°N due to convergence of the two plates. Further to the south (about 2°N), simple gravity models suggest that about 10 km of Caroline lithosphere lies beneath the present-day Pacific plate. Using a previously determined pole of rotation describing Caroline-Pacific relative motion (Weissel and Anderson, 1978), we grossly estimate the duration of the convergence between these two plates at about one million years. It is suggested that variation in the convergence rate along the plate boundary provides the primary control on the variation of structural deformation observed between provinces; however, favorable thermal conditions are factors that are considered. If the eastern boundary of the Caroline plate is a region of incipient though perhaps transient subduction, as we postulate, then the geophysical and geological evidence presented can constrain models on the initiation of subduction.

Description: Three arrays of ocean bottom seismographs have been deployed to study the seismicity at the northern end of the Juan de Fuca ridge system off western Canada. Nearly 100 events were located with estimated accuracies generally better than ±10 km, all lying on or near the en echalon ridge-transform fault plate boundaries as defined in this area by the magnetic anomalies, the seafloor morphology and by other geophysical data. The depths of 12 events were determined to lie between 2 and 6 km below the top of the crust. The seismograms exhibit clear P and S wave arrivals along with phases that involve P to S and sometimes S to P conversion probably at the base of the sediments beneath the instruments. The event magnitudes have been estimated from signal duration using four calibration events that were well recorded by a land station. The magnitude estimates permit the determination of rough magnitude-frequency of occurrence relations over the magnitude range of 1 to 3 that are in surprisingly good agreement with the recurrence relations for the area at larger magnitudes from 75 years of land station data. The mean P wave velocity in the uppermost mantle from the earthquake data recorded by the sea floor arrays is 7.6 km s-1 and the mean Vp/Vs ratio is 1.71 or a Poisson's ratio of 0.24.