Obtaining good sub-basalt seismic images is known to be problematic (Ziolkowski et al., 2003; White et al., 2003). Although the properties of basalts are quite different from those of most sediments Planke (1999) suggested that seismic energy is transmitted through basalt in much the same way as through sediments so the problem of seismic imaging through basalts amounts to the conventional task of separating primary energy from noise, even though the noise including multiples may be considerable. The physical properties of basalt are markedly different from those of the overlying and underlying sediments. Strong reflections due to high impedance contrasts at the top (and bottom) of the basalts leads to significant loss of transmitted seismic energy (Fruehn et al., 2001). Large variations of intrinsic properties along vertical cross-sections of basalt flows have been demonstrated and quantified by analyses of well-logs from wells penetrating successions of flood basalts (Planke, 1994) and from surface mapping (e.g., Self et al., 1998; Thordarson and Self, 1998). This causes the stratigraphic filtering effects of basalt successions to be more severe than that of sediments (Maresh and White, 2005; Shaw et al., 2004). Lateral variations in the thicknesses of sediments interbedded between basalt flows and in the thickness of the upper porous part of basalt flows have been demonstrated by detailed investigations in exposed flood basalts (Self et al., 1998; Thordarson and Self, 1998). The roughness of inter-beds causes 3D scattering of seismic energy, as demonstrated in studies comparing stratigraphic filtering and the effective quality factor, Q, of basalt successions (e.g., White et al., 2005; Shaw et al., 2005; Shaw et al., 2004). Taking these problems into consideration, experiments have been performed in the last decade using: longer offsets (both synthetic aperture, and longer streamers) to improve the signal-to-noise ratio and NMO resolution and to allow processing of post-critical reflections; larger energy sources to increase the general energy level; low-frequency tuning to allow for better penetration through basalts (characterised by low Q values); and shot-by-shot recording of the source signature and combination of OBS and seismic reflection data to improve velocity estimates. In one experiment all of the above-mentioned techniques were applied, providing considerable improvements in sub-basalt imaging relative to previous work (Spitzer and White, 2005; White et al., 2005). An other parameter for seismic acquisition is the orientation of the seismic line relative to the flow direction of the basalt flows (Reshef et al., 2003). However, poor effective transmission of seismic energy, scattering, strong multiple reflections, multiple mode conversions, and low-pass filtering of the energy that propagates through a layer of stacked basalt flows are still hampering routine imaging for petroleum exploration in sediment basins covered by basalts (Maresh and White, 2005). This was demonstrated by the UK164/07- 01 well where the base of a basaltic succession was found 700 m deeper than anticipated from interpretation of seismic reflection data (Archer et al., 2005). In order to obtain imaging quality and detail comparable to those obtained in other sedimentary basins, further improvements are necessary. The SeiFaBa Project (Seismic and petrophysical properties of Faroes Basalts), sponsored by the Sindri group, aims to create data-derived models for the propagation of seismic energy in basalt to provide a basis for better sub-basalt imaging. The project comprises drilling of the Glyvursnes-1 wells near Tórshavn on the Faroe Islands (Figure 1), core analysis for intrinsic physical parameters, recording of VSP and offset-VSP data in the Glyvursnes-1 and Vestmanna-1 wells, and surface seismic wide-angle and reflection data around the Glyvursnes- 1 and Vestmanna-1 wells (Japsen et al., 2005). At both sites these investigations of the elastic properties of basalts are made at a number of different scales. In this paper we present surface seismic reflection data from SeiFaBa experiment at Glyvursnes in the summer of 2003 illustrating that basalts can be imaged effectively using relatively small energy sources (250 g dynamite; 2.6-litre airgun cluster) and that stratigraphic details of flood-basalt constructions can be identified and characterized based on analysis of seismic data and then correlated to well data. We also demonstrate how different acquisition and processing techniques influence the effective frequency content of seismic reflection data and thus the effective propagation through and imaging of the basalts.
- Programområde 3: Energiressourcer