Summary In this paper the results of a tomographic analysis of a 3‐D wide‐angle seismic refraction data set acquired at the Valu Fa Ridge (VFR) in 1995 are presented. The VFR is an intermediate‐spreading ridge located in the southern Lau back‐arc basin in the southwest Pacific. The ridge comprises three morphological segments, the Southern, Central and Northern Valu Fa Ridges, separated by overlapping spreading centres (OSCs). Previous seismic experiments have identified a robust axial magmatic system beneath the central segment (CVFR) and the OSC with the northern segment (NVFR). The experiment described in this paper aimed to resolve details of the structure of this magma chamber and the adjacent post‐rift crust. A regularized inversion scheme that minimizes model roughness was applied to the first‐arrival traveltime picks made from the wide‐angle data. A quantitative approach for determining data uncertainties is described based on the signal‐to‐noise ratio of the arrivals. Several initial model assumptions were tested, including one with a thin melt lens, representing a seismic reflector identified in previous studies, explicitly included in the initial model. The inversion results suggest that crustal layer 2 exhibits northward thickening, which mirrors a similar northward thickening of the whole crust. In addition, local thinning of layer 2 is identified in the vicinity of the boundary between pre‐ and post‐rift crust, which is thought to represent thinning of the crust prior to the onset of rifting. Axial low‐velocity anomalies are identified in layer 2B/C and layer 3. The models are consistent with a continuous ∼ 6 km wide negative velocity anomaly in layer 3 with an amplitude of ∼ 0.7–0.9 km s−1 relative to off‐axis post‐rift crust. This anomaly is consistent with the presence of an axial mush zone comprising a small percentage (< 1 per cent) of partial melt. The negative velocity anomaly in layer 2B/C is modelled with its largest amplitude (∼0.5 km s−1) beneath the northern OSC. Possible origins for this anomaly include locally thicker crust or locally higher porosity near the OSC, or a high‐temperature anomaly associated with the axial magmatic system.

No abstract

No abstract

Summary Over the last 20–30 yr numerous seismic images of the Earth's crust have revealed details of its gross structure, including intra‐crustal layering, the geometry of that layering and its composition. As more and higher quality studies are undertaken it is becoming apparent that identified structures have a greater degree of 3‐D variability than first anticipated. Thus, the methodology of crustal imaging by seismic means has also developed into the third dimension with a tomographic approach now being widely adopted, particularly so in the marine environment. Such surveys not only focus on mapping the finer scale 3‐D structural variability, they also aim to achieve sufficient density of azimuthal coverage and resolution to address preferential orientation patterns of features such as porosity, fracturing and faulting. Recent developments in technology, and consequently cheaper construction and deployment costs of instruments, have resulted in an expansion in the number of instruments available in ocean‐bottom seismometer pools. Consequently, individual experiments are being designed to accommodate the maximum number of instruments available and this, coupled with dense grids of shot profiles, significantly impacts on survey cost. In this paper we consider a variety of approaches to achieving the best resolution of detail for minimal associated cost of acquisition, and for instrument pools of various sizes. A number of different geometries are compared, including example grid designs in current use. Comparison of resolution tests and relative costings for a range of acquisition geometries suggest that, if instrument numbers and/or funds are limited, the most cost effective ways of achieving the desired target resolution may be by (1) shooting additional shot profiles at the expense of deploying more instruments and (2) multiple, overlapping deployments of a small geometry, tailored in shape to the target structure and depth.

No abstract