Determination of the volcanostratigraphy of oceanic crust formed at superfast spreading ridge: Electrofacies analyses of ODP/IODP Hole 1256D

Abstract

The objective of this study is to construct a representative volcanostratigraphy of Ocean Drilling Program Hole 1256D, the first complete penetration of intact upper oceanic crust formed at a superfast spreading rate. An accurate knowledge of the volcanostratigraphy is vital to understand processes of crustal construction and submarine magmatism and to estimate chemical exchange with seawater, but this is rarely achieved due to very low recovery rates in most basement holes. We used two approaches to determine the rock types that form the wall rocks in the basement sections of Hole 1256D: (1) user guided interpretations of electrofacies acquired by imaging tools combined with other wireline tools; and (2) the use of an artificial neutral network to objectively classify the responses of all available logging information. Great availability of formation microscanner (FMS) images provided superior coverage of the borehole wall compared to previous attempts at core‐log integration. This has resulted in more confident and detailed lithologic classifications, such as with the distinction between pillows and different styles of breciation. Ten lithology types are suggested for a volcanostratigraphy model: massive flows, ponded lava, fractured massive flows, fragmented flows, thin flows or thick pillows, pillows, breccias, dikes in dike complex, isolated dikes, and gabbros. Three major lithology types in the extrusive section are massive flows (both massive and fragmented massive flow, 22%), fragmented flows (32%), and breccias (19%). Pillow lavas make up only 1.9% of the volcanic section and are confined to a 100 m interval. Below the extrusive section, subvertical contacts interpreted to be dike margins are typically observed every 1 to 2 m with brecciated zones along the contacts. The dikes dip steeply to the northeast indicating slight rotation away from the ridge axis. We used an artificial neural network (ANN) approach to determine a quantitative lithostratigraphy. The ANN is most strongly influenced by porosity and alteration degrees and the resulting stratigraphy most closely resembles the above classifications when clustered by FMS texture as opposed to lithologic interpretation. The ANN thus provides a porosity‐based stratigraphy of the basement rather than the traditional lithology‐based stratigraphy.