Seismic methods for uranium exploration: an overview of EXTECH IV seismic studies at the McArthur River mining camp, Athabasca Basin, Saskatchewan

Transcription

1 Seismic methods for uranium exploration: an overview of EXTECH IV seismic studies at the McArthur River mining camp, Athabasca Basin, Saskatchewan D.J. White 1, Z. Hajnal 2, B. Roberts 1, I. Györfi 2, B. Reilkoff 2, G. Bellefleur 1, C. Mueller 1, S. Woelz 1, C.J. Mwenifumbo 3, E. Takács 2, D.R. Schmitt 4, D. Brisbin 5, C.W. Jefferson 3, R. Koch 6, B. Powell 5, and I.R. Annesley 7 White, D.J., Hajnal, Z., Roberts, B., Györfi, I., Reilkoff, B., Bellefleur, G., Mueller, C., Woelz, S., Mwenifumbo, C.J., Takács, E., Schmitt, D.R., Brisbin, D., Jefferson, C.W., Koch, R., Powell, B., and Annesley, I.R., 27: Seismic methods for uranium exploration: an overview of EXTECH IV seismic studies at the McArthur River mining camp, Athabasca Basin, Saskatchewan; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588, p (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Abstract: Seismic-reflection data and a vertical seismic profile were acquired in the vicinity of the McArthur River mining camp. These data are interpreted with the aid of in situ geophysical and geological logs and rock-property measurements, which indicate that reflectivity within the basin-fill strata is controlled largely by porosity variations (Φ = 11%) that are attributed primarily to zones of silicification (postdepositional hydrothermal horizons), but also to grain-size lithological variations. The reflection data clearly image the unconformity zone and associated fault offsets including the P2 mineralized fault zone. A prominent shallow- to moderately dipping zone of reflections that extends downward from the surface location of the P2 fault is interpreted as a major crustal shear zone that partially controlled the locus of high-grade uranium ore deposition. The seismic techniques have demonstrated their utility in defining some of the key geological variations that are relevant to identification of prospective ores zones. Résumé : Des données de sismique-réflexion et un profil sismique vertical ont été acquis à proximité du camp minier de McArthur River. Ces données sont interprétées à l aide de diagraphies géophysiques et géologiques in situ et de mesures des propriétés des roches, qui indiquent que la réflectivité dans les strates de remplissage du bassin est en grande partie déterminée par les variations de la porosité (Φ = -11%) associées principalement aux zones de silicification (horizons hydrothermaux postsédimentaires) mais aussi aux variations granulométriques des lithologies. Les données de sismique-réflexion montrent clairement la zone de discordance et les décalages produits par des failles associées, incluant la zone de failles minéralisée P2. Une zone de réflexion bien visible, de pendage faible à modéré, qui s étend en profondeur depuis l expression en surface de la faille P2, est interprétée comme étant une zone de cisaillement crustale majeure qui a partiellement contrôlé la mise en place du minerai d uranium à forte teneur. L utilité des méthodes sismiques a été démontrée pour la définition de certaines variations géologiques pertinentes lors de l identification des zones minéralisées prometteuses. 1 Geological Survey of Canada, 615 Booth Street, Ottawa, Ontario K1A E9 2 Department of Geological Sciences, University of Saskatchewan, Saskatoon, Saskatchewan S7N 5E2 3 Geological Survey of Canada, 61 Booth Street, Ottawa, Ontario K1A E8 4 Institute for Geophysical Research, Physics Department, University of Alberta, Edmonton, Alberta T6G 2J1 5 Cameco Corporation, th Street West, Saskatoon, Saskatchewan S7M 1J3 6 AREVA Resources Canada Inc., P.O. Box 924, th Street West, Saskatoon, Saskatchewan S7K 3X5 7 Saskatchewan Research Council, Innovation Boulevard, Saskatoon, Saskatchewan S7N 2X8 363

2 GSC Bulletin 588 INTRODUCTION A comprehensive seismic study was conducted as part of the EXTECH IV Athabasca Multidisciplinary Uranium Studies Project (Jefferson and Delaney, 21). The two overall goals of the seismic program were to test the effectiveness of seismic techniques as tools for uranium exploration, and to contribute to the four-dimensional geoscience framework for uranium exploration within the deeper recesses of the Athabasca Basin (Fig. 1). Here, and in companion papers (Györfi et al., 27; Hajnal et al., 27), results from the various seismic investigations are presented. The specific objectives of the seismic program were: 1) define the subsurface stratigraphy of the sedimentary rocks within the basin, 2) provide a detailed image of the basement unconformity that hosts the majority of the uranium ore deposits, 3) characterize the basement unconformity using seismic attributes and identify attributes that define zones of mineralization, 4) locate and image faults that have been instrumental in ore deposition, 5) determine the seismic signature of a known ore deposit, and 6) define the regional basement structure underlying the basin including faults. These objectives were addressed by applying a variety of seismic techniques (see Fig. 2 (colour folio, p. 61); 3 for locations). High-resolution seismic methods were utilized to achieve objectives 1 to 5 including 8 km of high-resolution 2-D seismic profiling along lines 12 and 14 (Fig. 3)(see also Györfi et al., 27); a limited, 3-D high-resolution survey; high-frequency, near-offset and far-offset three-component vertical seismic profiles (VSPs) using borehole MAC-218; and 3-D vertical seismic profiling. The 2-D and 3-D surveys included limited three-component recording. To address objective 6, 34 km of deep-sounding seismic-reflection profiles were acquired along two regional lines (A-A' and B-B' of Fig. 2; see also Hajnal et al. (27)). Györfi et al. (27) presented a detailed interpretation of the 2-D high-resolution seismic profiles through the mining camp, and Hajnal et al. (27) provided an interpretation of the regional seismic profiles in terms of the tectonic framework of the western Trans-Hudson Orogen. Here, the present authors summarize results for the entire seismic program, but given the complementary nature of the aforementioned companion papers, this paper focuses on topics not described there, including results from downhole vertical seismic profiling and the high-resolution 3-D survey. Prior to presentation of the results from the field program, this paper provides a discussion of expected reflectivity characteristics for the McArthur River geological setting based on geophysical logging, rock-property measurements, and numerical simulation of the reflection response for a model of the orebody. All of these results are considered with a view to effective use of seismic methods for further exploration within the Athabasca Basin and other analogous basins (e.g. Thelon and Hornby basins). GEOLOGICAL SETTING The McArthur River mine in the Read Lake study area is located in northern Saskatchewan (Fig. 1) in the eastern part of the Athabasca Basin at the transition between the basement Wollaston and Mudjatik domains of the Archean-Paleoproterozoic Hearne Province. In the area of investigation, the western Wollaston and transitional Wollaston-Mudjatik basement lithostructural domains have been multiply deformed, interleaved with Archean basement, and metamorphosed (Lewry and Sibbald, 198; Portella and Annesley, 2a, b; Tran, 21) within the southeastern Hearne Province margin during ca. 1.8 Ga continent-continent collision recorded by the Trans-Hudson Orogen (Hoffman, 199). In the study area, the Athabasca Group sandstone basin fill is 4 6 m thick and comprises only the Bird (MFb), Collins (MFc), and Dunlop (MFd) members of the Manitou Falls Formation (Fig. 1; Ramaekers (199)) that overlie the Read Formation (RD; former MFa). Although recognizing that proposed stratigraphic revisions by Yeo et al. (27) are valuable, this paper continues the working usage for this project (Bernier et al., 21; Yeo et al., 27). The unconformity-associated uranium ore deposits in the basin are well documented (e.g. Hoeve and Quirt, 1984; Sibbald, 1986). The McArthur River mine site is located along the northeast-trending P2 fault zone. At least four super-high-grade uraninite bodies ( t U in Proven and Probable Reserves at 17.96% U; Thomas et al. (2)) are located at a depth of about 55 m (Jamieson and Spross, 2), where the unconformity underlying the Athabasca Group and the associated paleoweathered zone, are offset up to 8 m by moderately dipping reverse faults (e.g. P2 fault of Fig. 4) that, in plan, obliquely intersect a conductive zone of graphite-rich, pelitic gneiss of the Wollaston Group (McGill et al., 1993). PREVIOUS SEISMIC WORK Seismic methods have not been extensively tested in the Athabasca Basin, although the reconnaissance refraction surveys of Hobson and MacAuley (1969) and Overton (1977) regionally mapped the depth of the sandstone-basement contact throughout the basin. Subsequent attempts to use this technique are not documented in the literature. Several rudimentary reflection experiments (Hajnal and Reilkoff, 198; Scott, 1983; Fouques et al., 1986) met with only marginal success. More recent seismic-reflection investigations, through collaboration of LITHOPROBE and a consortium of mining companies, implemented the latest technology with greater success (Hajnal et al., 1997). Those successes were largely responsible for initiation of the new work described in this study. 364

3 Figure 1. Regional geology of the Athabasca Basin region in northern Saskatchewan and Alberta. The approximate location of the seismic study area and transects at the McArthur River mining camp (Fig. 2) are outlined by the rectangle. Previous seismic transects in the Athabasca Basin are shown approximately in the Dawn Lake and Shea Creek areas. Deposits and significant occurrences are listed in Jefferson et al. (27). Basement geology is after Card et al. (27a, b), Portella and Annesley (2a, b), and Thomas et al. (22). Athabasca Basin geology is after Ramaekers et al. (27). Strata in the McArthur River area comprise two formations of sequence 2. The basal unit there is the Read Formation (formerly MFa). The overlying Manitou Falls Formation in the McArthur River area comprises the Bird Member (MFb), Collins Member (MFc), and Dunlop Member (MFd). Generalized fault zones (these include multiple ductile movements before deposition of Athabasca Group and brittle transcurrent and dip-slip movements during and after deposition) are: A = Allan; BB = Black Bay; BL = Black Lake; BR = Beatty River, CB = Cable Bay, D = Dufferin; GR = Grease River; H = Harrison; NF = Needle Falls; PL = Parker Lake; P2 = P2 fault at McArthur River; R = Robillard; T = Tabbernor; VR = Virgin River shear zone (Dufferin is one named fault of many in VR); Y = Yaworski. Contents Author Index D.J. White et al. 365

4 GSC Bulletin 588 NW SE Projection line 15 P2 trend RL D-1 MAC MAC MAC RL-68 MAC MAC D-2 85 McArthur River mine 9 2 km SEISMIC REFLECTIVITY OF THE BASIN ROCKS N Easting (km) In this section, an attempt is made to quantify the effects of geological variations in controlling the observed reflectivity within both the sedimentary and crystalline rocks of the McArthur River area. This is done using in situ geophysical logging measurements, laboratory measurements on rock cores, estimates of mineral composition, measured mineral properties, and the results of vertical seismic profiling. Few borehole data are available from the gneissic basement rocks because drilling was stopped soon after ore was intersected, and the basement part of drillholes is normally filled with 6 MC Figure 3. Location map for the 2-D high-resolution seismic lines 12 and 14. Also shown are the projection line for the line 12 seismic image (Fig. 2), the approximate locations of the McArthur River ore zones (numbered ellipses) and mineralized zone (dashed line labelled P2 trend ), and the locations of boreholes referred to in the text (circles with Xs). A VSP was acquired in borehole MAC-218. The black rectangle surrounding the McArthur River mining camp designates the same region as the black rectangle in Figure 2. Numbers along lines 12 and 14 are common depth point station numbers. 9 Northing (km) Sandstone Conglomerate Unconformity Basement gneiss Sandstone Conglomerate Unconformity 5 m Basement gneiss P2 fault High-grade ore Low-grade ore ~55 m Figure 4. A schematic cross-section of the McArthur River orebody (after Jamieson and Spross, 2), used as the basis for the model in Figure 1. The unconformity occurs at about 55 m depth, and the model is plotted without vertical exaggeration. concrete. To compensate for this lack of information about the basement rocks, the authors estimate the effects of mineralogical composition using known mineral properties. In the McArthur River Read Lake area (Fig. 1), the Athabasca Basin preserves most of the Manitou Falls and Read formations, an overall upward-fining succession of quartzose sandstone, conglomerate, and mudstone of fluviatile origin that rests unconformably on intermediate metamorphic grade interleaved crystalline and supracrustal basement gneiss. The unconformity surface between the sedimentary strata and gneiss was expected to be the primary reflector in the study area. Additional zones and variations of reflectivity were expected due to differing rock properties within the stratigraphic column, across and along the unconformity and paleoweathered zone, and among the numerous basement rock units. Documented geological variations (McGill et al., 1993; Mwenifumbo et al., 2, 21, 22, 24; Jefferson et al., 27) include 1) a discontinuous lower conglomerate unit immediately above the basal unconformity, 2) intensely fractured zones within the basal strata, 3) zones of silicification within the strata, 4) compositional variations of the unconformably underlying Hudsonian gneiss, and 5) the presence of a variably paleoweathered layer at the top of the gneiss. Györfi et al. (27) summarized the geological attributes of some of these features. 366

5 D.J. White et al. The stratigraphic column Strata of the Manitou Falls and Read formations have a relatively uniform bulk siliciclastic mineralogical composition and thus might be considered to have generally weak reflectivity; however, compressional-wave velocity (V p ) varies with porosity and volume-fraction clay in detrital silicate rocks (Tosaya and Nur, 1982; Han et al., 1986; Eberhart-Phillips et al., 1989). For volume-fraction clay contents of more than a few per cent, V p is generally more sensitive to porosity than to clay content by a factor of three to four (e.g. Han et al., 1986), but very low volume-fraction clay contents (<.3) can significantly affect the velocity (by as much as 7%). The chemical composition of the dominant clay mineral is generally a second-order factor in affecting V p (Tosaya and Nur, 1982). Figure 5 portrays compressional-wave velocity V p, density (ρ), and acoustic impedance (Z= ρ V p ) versus porosity (Φ) for a limited set of core samples that are predominantly from the sedimentary column at McArthur River. The V p and porosity values (Fig. 5a) for the sedimentary samples lie predominantly within the ranges of km/s and 1%, respectively, comparable with values from elsewhere in the basin where V p of km/s and 5 1% porosity were measured (Hajnal et al., 1983). For comparison, a general empirical relationship between V p, porosity, and clay content for sandstone (Eberhart-Phillips et al., 1989) is shown in Figure 5a for volume-fraction clay contents of. and.3. This empirical relationship accounts for the increased sensitivity of V p for low clay contents. For porosities ranging from to 1%, there is a clear negative trend in the associated values of V p, density (Fig. 5b), and impedance (Fig. 5c) that decrease by about 13%, 5%, and 22%, respectively. As can be seen by inspection of the solid lines in Figure 5a, a maximum decrease in V p of about 6% (i.e. the span between the solid lines for a constant porosity value) that is potentially attributable to variations in clay content is much smaller than the decrease in V p of about 13% due to porosity variations. The broad scatter in the values of V p may in part be due to variability in clay content, variations in the mineralogy of the sandstone, and/or alteration. The group of outliers that have V p more than 5.6 km/s and densities exceeding 2.8 g/cm 3 are basement (calc-silicate gneiss) samples. The lowest values of V p are likely due to the low pressures at which the measurements were made, allowing crack-like pores to remain partially open. Such porosity contributes little to the bulk porosity (compare reduced scatter in density values in Fig. 5b), but is very effective in reducing V p. The variations of density (ρ) and acoustic impedance as a function of porosity and volume-fraction clay content are portrayed in Figures 5b and 5c. The measured impedance values (Fig. 5c) ranging from 1 x 1 6 kg s -1 m -2 to15x1 6 kg s -1 m -2 are generally consistent with the range of impedances determined by in situ logging (Table 1; Fig. 6). Larger scale fractures that are not sampled by the rock cores may increase in situ porosity further. The dominance of porosity versus clay content effects is accentuated for density (Fig. 5b) where associated density reductions are about 5 6% for porosity as compared to about 1 2% for clay content. Acoustic impedance variations of up to 22% are associated with the observed range of porosities, whereas the variations attributable to clay content are less than 1%. Thus, based on the rock-property measurements the authors conclude that in the McArthur River area where volume-fraction clay contents are generally low (.1.3), porosity-related variations in acoustic impedance will be dominant over clay-related variations by a factor of two to three; however, in other parts of the basin where clay content is much higher (e.g. classic unconformity-type uranium deposits such a Cigar Lake or Midwest; Fig. 1), the two factors may be comparable. For samples obtained near the Midwest uranium deposit (Fig. 1), Hajnal et al. (1983) reported V p decreasing from 4.8 km/s to 4.3 km/s (or 1%) for volume-fraction clay increasing from about. to.4. Downhole geophysical logs (Mwenifumbo et al. (24); summarized in Table 1, Fig. 6) document relatively large changes in in situ density, V p, and acoustic impedance within the Manitou Falls Read formation strata that show a close correlation with resistivity variations (e.g. Fig. 7, see colour folio, p. 62). This correlation suggests that the observed variations are largely porosity controlled (Mwenifumbo et al., 24), corroborating the same conclusion deduced from the core measurements. Porosity, in turn, is primarily controlled by silicification and desilicification, fracturing, and to a lesser extent, grain size (Mwenifumbo et al., 27). Fracturing is most apparent in steep, narrow fault zones that transect the sandstone column at high angles (cf. Fig. 3 in McGill et al., 1993) and broader damage zones at the base of the section in the vicinity of basement faults. Reverse fault splays that flatten into bedding-parallel slip and kink folds at their terminations may also result in bedding-parallel fracturing or diffuse bodies of fracturing higher in the stratigraphy (cf. Tourigny et al., 22). Silicification, in contrast, trends approximately parallel to bedding on mesoscopic scales, and probably feathers out laterally. Exploration core logging (McGill et al., 1993; Thomas et al., 2) and gravity modelling (Thomas and Wood, 27) indicated that silicified zones extend laterally for hundreds of metres, and thus potentially represent prominent reflectors within the sandstone column. Large acoustic impedance variations and reflection coefficients (R) are associated with both silicification (maximum of 26% increase in Z due to silicification with a corresponding value of R =.11) and fracturing (15% reduction in Z due to fracturing with a corresponding value of R = -.8) according to Mwenifumbo et al. (24, 26). A pertinent question for seismic interpretation is whether the reflectivity portrays hydrothermal and diagenetic processes only, or whether the stratigraphic units within the basin strata are recognized. The effects of silicificationrelated porosity variations has been clearly demonstrated, but close inspection of the V p and density logs in comparison to the various lithostratigraphic parameters for MAC-218 (Fig. 7) suggests possible correlations with stratigraphic horizons. For instance, the increase in V p and density at about 32 m (dashed line in Fig. 7), which is attributed primarily to increased silicification, also correlates with a subunit boundary within MFb where the content of conglomerate increases sharply. Similarly, at the contact between MFb1 and RD4 367

6 GSC Bulletin 588 a) V (km/s) p MAC-257, MAC-26 V versus porosity p Porosity (%) MAC-26 MAC-257 E-P estimate: no clay E-P estimate: 3% clay Linear (E-P estimate: no clay) Linear (E-P estimate: 3% clay) b) c) Wet density versus porosity Density (g/cm 3 ) MAC-26 MAC-257 H 2 adj. density 3% clay density Linear (H 2 adj. density) Linear (3% clay density) Impedance Porosity versus impedance (wet V, wet density) p MAC-26 MAC-257 No clay 3% clay Linear (no clay) Linear (3% clay) Porosity (%) Porosity (%) Figure 5. Compressional-wave velocity (V p ), porosity (Φ), density (ρ), and acoustic impedance determined by laboratory measurements on core samples from boreholes MAC-257 and MAC-26. a) Compressional-wave velocity versus porosity: V p measurements were made on water-saturated cores with a uniaxial pressure of about 3 MPa applied to the rock cylinder. The solid lines are V p values estimated from measured porosity using an empirical relationship for sandstones (V p = Φ-1.73C 1/ (P e e -16.7Pe )) from Eberhart-Phillips et al. (1989)), where Φ is fractional porosity, C is fractional clay content, P e is effective pressure, and V p is expressed in km/s. The lines correspond to volume-fraction clay contents of. and.3 that are representative of the range of clay contents at McArthur River. b) Density versus porosity: densities were determined using core volumes and weights measured after vacuum oven drying of the cores for 48 h. The densities shown in the figure are estimated densities for water-saturated cores where the measured dry densities have been adjusted for water saturation using the measured porosities. The solid lines indicate densities estimated from porosity for volume-fraction clay contents of. and.3 calculated using ρ =(1-Φ) +Φ W where ρ = ρ M (1-C) + ρ C C, for ρ M = unaltered rock matrix density = 2.67 g cm -3, ρ C = clay density = 1.44 g cm -3, ρ W = water density = 1. g cm -3, and C =. and C =.3. c) Acoustic impedance (Z=ρ V p with units of 1 6 kg s -1 m -2 ) versus porosity for water-saturated conditions. The solid lines indicate impedances estimated using the results from Figures 5a, b for volume-fraction clay contents of. and.3. H 2 O adj. density = water-adjusted density. 368

7 D.J. White et al. Table 1. Mean values for sandstone, silicified sandstone, and basement rocks from in situ downhole geophysical logging from Mwenifumbo et al. (2, 21, 24). Standard deviations (in parentheses) are provided when available. Lithological unit Mean V p (km s -1 ) Mean density (g cm -3 ) Acoustic impedance (x1 6 kg s -1 m -2 ) Sandstone Sandstone Silicified sandstone Basement Basal conglomerate 5.66 (.18) 2.59 (.4) 14.7 p Drillhole In situ: Psammitic gneiss Quartzite Pegmatite Semipelitic gneiss Garnet-cordierite pelitic Gneiss Sandstone 1 Sandstone 2 Silicified sandstone Basal conglomerate Compositional estimates: Quartzite Psammitic gneiss Pelitic gneiss Estimates In situ Figure 6. Velocity (V p ) versus density (ρ) for 1) in situ measurements within the sedimentary column (black symbols; Mwenifumbo et al. (24)), 2) mean values determined for populations of the former (sandstone 1, sandstone 2, silicified sandstone, basal conglomerate described in Mwenifumbo et al. (24)), 3) mean values determined from in situ measurements of basement units (quartzite, psammitic gneiss, semipelitic gneiss, garnet-cordierite pelitic gneiss, and pegmatite) within a single borehole penetrating the basement, and 4) estimated for various basement rock types (quartzite, psammitic gneiss, pelitic gneiss). Iso-impedance curves (Z with units of 16 kg s -1 m -2 ) are also shown. The symbols for the basement units are grouped by the ellipses according to whether they are in situ measurements or estimates from mineralogical composition (solid colours). Impedance values for the high-grade uranium ore zones range from 32 x 1 6 kg s -1 m -2 to 5 x 1 6 kg s -1 m -2 using the in situ logging velocities from Mwenifumbo et al. (22) and an assumed density of 1. g cm -3. (at ~38 m depth in Fig. 7), there is an abrupt increase in density and the content of conglomerate decreases sharply. The control of grain size on porosity has been noted (Mwenifumbo et al., 27). Thus, it is likely that porosity (the dominant factor determining acoustic impedance in the strata) is at least partly controlled by stratigraphy, either directly (by grain size), or indirectly through associated permeability characteristics that controlled fluid flow, which ultimately led to zones of silicification and desilicification. Although the main stratigraphic units show little clear relationship with acoustic properties, detailed visual comparison (Fig. 7) shows a consistent inverse relationship between grain size (gamma-ray and maximum grain-size logs) and acoustic velocity (as logged by density and velocity) on a scale of 5 1 m. This decametre-scale relationship is superimposed on broader fluctuations in each parameter that do not correlate. The interpretation is that coarser grained beds have higher porosity than immediately adjacent, finer grained beds, consistent with qualitative visual examination of drill core. This relationship holds for cobble conglomerate versus adjacent granule-rich sandstone of RD and MFb, as well as for granule-rich sandstone versus adjacent fine-grained sandstone. Geophysical logs for several boreholes within the McArthur River camp are shown in Figure 8 and demonstrate the variability in reflectivity within the sandstone column. In all of the borehole logs, a zone of high acoustic impedance is observed in the lower portion of the sandstone (at ~3 35 m), corresponding to a zone of silicification. The vertical abruptness of the associated acoustic impedance change (Fig. 8) determines whether prominent seismic reflections are observed (e.g. boreholes MAC-218, MAC-257, MAC-258, and MAC-259 reported here, and MAC-218 and RL-88 reported by Mwenifumbo et al. (24)) or not (boreholes MAC-197 and RL-92 reported by Mwenifumbo et al. (24) show gradational density increases at gradual silicification fronts). Prominent reflections are also observed from thin zones with anomalously low acoustic impedance that are likely fractures (e.g. indicated by label Fr in Fig. 8). The lateral extent of the fracture zones is uncertain, but the present authors suspect that in most cases they are local features. Basement rocks Composition of the major basement lithological units is variable and includes pelitic gneiss (>18% garnet+biotite±cordierite, with biotite and garnet ranging from 15 25% and 3 1%, respectively), psammitic gneiss (<9% SiO 2 and up to 2% sillimanite), quartzite (>9% SiO 2 ), and granite. Rock-property measurements on basement rocks from the McArthur River region are limited, as are in situ geophysical logs that penetrate the basement (exceptions are RL-92 (Fig. 6), MC-265, and MAC-257 (Fig. 8); see also Mwenifumbo et al., 24). The values of in situ acoustic impedance for the basement in RL-88 (Fig. 6) and for MC-265 provided by Mwenifumbo et al. ((24); Table 1) are very low compared to those of MAC-257 (Fig. 8b), and thus are interpreted as altered by paleoweathering and/or hydrothermal alteration. 369

8 GSC Bulletin 588 a) MAC-218 b) MAC-257. VELOCITY (m s ) -1 ACOUSTIC IMPEDANCE RC SERIES (kg s m ) -1-2 RAW SYNTHETIC WELL TOPS MFd3 VELOCITY (m s ) -1 ACOUSTIC IMPEDANCE RC SERIES RAW SYNTHETIC (kg s m ) -1-2 WELL TOPS MFd MFd2 MFd1 MFc4 MFc3 MFc2 MFc1 MFb3 MFb2 MFb1 RD MFc MFb RD.2 5 RD1 Basal conglomerate.2 5 RDs Uc Reg.25 TIME (s) 6 DEPTH (m).25 TIME (s) 6 DEPTH (m) VELOCITY (m s ) MAC ACOUSTIC IMPEDANCE RC SERIES RAW SYNTHETIC WELL TOPS VELOCITY (m s ) MAC ACOUSTIC IMPEDANCE RC SERIES RAW SYNTHETIC WELL TOPS Fr Fr c) (kg s m ) -1-2 d). (kg s m ) MFd.5 Fr 1.5 Fr MFc.1 MFb Fr 2 3 MFc MFb RD RD Uc 5 Uc.25 TIME (s) 6 DEPTH (m).25 TIME 6 (s) DEPTH (m) Figure 8. Compressional-wave velocity, acoustic impedance (Z with units of 16 kg s -1 m -2 ), and reflectivity series (RC) determined from borehole sonic logs, and calculated seismograms: a) MAC-218, b) MAC-257, c) MAC-258, and d) MAC-259. The seismograms are determined by convolution of the reflectivity series and an estimated source wavelet that is shown at the top of each panel of raw synthetic seismic traces. Fr = fracture zone. Also indicated are tops of the various members of the Manitou Falls Formation (MFb, MFc, MFd), Read Formation (RD, Rds), the unconformity (Uc), and regolith (Reg). Numbers following RD, MFb, MFc, and MFd are seismically differentiated subunits within these lithostratigraphic units defined by Yeo et al. (27). 37

9 D.J. White et al. Acoustic impedances have also been estimated for the unaltered basement rock types by using the single mineral properties in Table 2 to calculate values for isotropic mineral aggregates (Table 3). The V p estimates were obtained using a simple time-average equation (Wyllie et al., 1956; Sheriff and Geldart, 1995, p. 117) and aggregate densities are calculated as the volume-fraction weighted mean of the constituent mineral densities. Perusing the acoustic impedance values in Table 2, it is clear that acoustic impedance in unaltered basement rocks will be strongly affected by the percentage of the aluminosilicate minerals cordierite, sillimanite, and garnet as the acoustic impedance values of these minerals are about 5% (cordierite) to about 1% (sillimanite and garnet) greater than quartz and plagioclase. As displayed in Figure 6, the impedance values of the unaltered basement rocks are generally higher (Z>16 x 1 6 kg s -1 m -2 ) and show less internal variation (<12%) than the overlying strata where Z<15 x 1 6 kg s -1 m -2 and variations of up to 3% are observed; however, altered basement rocks (Fig. 6) have lower impedance values that are similar to the silicified basal portion of the stratigraphic column and increase the expected variability of impedance within the basement. Given the small range in impedance values for the different basement rock types, it is unlikely that different basement domains can be distinguished based uniquely on reflection strength. Variability at the unconformity Reflection coefficients of R = are expected for a sandstone-to-basement transition implying that a strong reflection should be observed; however, variability in many attributes of the unconformity should significantly affect the associated reflectivity (see summary in Fig. 9). Highly silicified conglomeratic or sandstone subunits at the base of Read Formation have a higher impedance than the balance of the overlying sandstone (Table 2; Fig. 6), and thus where present may reduce the reflection strength at the unconformity to R = Silicified conglomerate could result in an additional reflection at the contact with the overlying sandstone (R =.5.15 for sandstone and silicified sandstone). Conversely, fracturing and desilicification within the lowermost Read Formation beds and along the unconformity has also been observed in a number of boreholes and is accompanied by a corresponding reduction in V p and to a lesser extent, in density. This constitutes a 3 5 m thick, low-velocity zone at the base of the sandstone column with significant lateral continuity that potentially may cause a negative polarity reflection underlain by a very strong positive polarity reflection (Fig. 9). The presence of a regolith should result in reduced velocities in the basement and hence reduced reflection strength in association with the unconformity. Logging shows that the thickness of the regolith is highly variable. A thicker regolith results in a broader depth range over which the velocities increase toward that of the fresh basement rocks, which too should result in reduced reflection strength at the unconformity zone. Summary Internal reflectivity of the sandstone will generally be controlled by porosity variations. Reflectivity should generally be weak (R <.5), except where zones of contrasting porosity are juxtaposed, for example where fracture zones (R = -.8; very local effect) and bedding-parallel silicification (R =.11) occur. The sandstone-basement contact should generally be a very strong reflector (R =.27.31), but the reflection strength will be reduced if the overlying sandstone is silicified (R =.11.16) and/or if a pronounced regolith is present. It will also be reduced in strength if silicification is intense through the base of the Manitou Falls Formation strata directly into fresh and/or silicified basement (R =.6.11). Extensive fracturing at the base of the sandstone will result in reflections from the top (R = -.11) and bottom (basement, R =.25) of the resultant low-velocity zone. REFLECTIVITY OF THE OREBODY Estimated impedance values for the uranium ore (Z = 32 x 1 6 kg s -1 m -2 to 5 x 1 6 kg s -1 m -2 based on the logging results of Mwenifumbo et al. (22)) are very high relative to the host rocks and thus have the potential to produce strong reflections; however, the size of the orebodies is ultimately the factor that determines whether they can be detected seismically. To investigate the expected seismic response of an orebody, the vertical-incidence seismic-reflection response of a representative uranium orebody was simulated. The 2-D model (Fig. 1) is based on a cross-section of the McArthur River orebody (Fig. 4) from Jamieson and Spross (2). The seismic response as recorded on vertical component (Fig. 11a, b, c, see colour folio, p. 63) and horizontal component (Fig. 11d, e, f) geophones is presented. Although only the vertical component response is directly comparable to the data from the multichannel, vertical-component seismic profile acquired in this study, the present authors also present the results for the horizontal component with a view to assessing Table 2. Physical properties of relevant minerals. Mineral Mean V p (km s -1 ) Mean density (g cm - 3 ) Mean V s (km s -1 ) Acoustic impedance (x1 6 kg s -1 m -2 ) Pressure (MPa) Source Garnet Ji et al. (22) Cordierite Toohill et al. (1999) Sillimanite Ji et al. (22) Biotite Lebedev et al. (1982) Quartz Ji et al. (22); 2 Lebedev et al. (1982) Plagioclase Ji et al. (22) = no data, V p = Compressional-wave velocity, V s = shear-wave velocity 371

10 GSC Bulletin 588 Table 3a. Physical properties of basement rocks estimated from mineral compositions using the properties in Table 2. The gneiss units are composed of the indicated mineral fractions with quartz constituting the remainder. Rock Quartzite 1% quartz Psammitic gneiss 2% sillimanite Pelitic gneiss 3% garnet, 15% biotite Pelitic gneiss 1% garnet, 25% biotite Complex unconformity Simple unconformity Sandstone column 8. ss/regolith/bsmt 7. ss/fr ss/bsmt 6. ss/fangl/bsmt 5. sil ss/bsmt 4. ss/bsmt 3. silicified ss 2. fractures 1. sandstone Mean V p (km s -1 ) Mean density (g cm -3 ) Acoustic impedance (x 1 6 kg s -1 m -2 ) Table 3b. Physical properties of basement rocks estimated from sonic and density logging in borehole MC-265. Properties are determined using a 5 cm moving average over the depth interval of m depth. Rock Mean V p (km s -1 ) Mean density (g cm -3 ) Acoustic impedance (x1 6 kg s -1 m -2 ) Quartzite Psammite Garnet cordierite pelite Semipelitic gneiss Pegmatite R (reflection coefficient) Figure 9. Summary of predicted reflection coefficients for various geological scenarios. See text for discussion. The values are based on the assumptions that transitions are abrupt, there are no thin layer effects, and there is no attenuation. Overlapping arrows are designed to indicate the expected variability in reflection coefficients for the associated geological scenario described in the text. Two arrows for a given scenario indicate that two distinct reflections may occur. Abbreviations: ss = sandstone, fr = fractured, fangl = fanglomerate, sil = silicified, bsmt = basement. the potential utility of horizontal component geophones in detecting ore zones. The following results are noteworthy. 1) The sandstone-basement interface results in a large amplitude, laterally continuous reflection on the vertical component (Fig. 11a, b), whereas it is generally invisible on the horizontal component (Fig. 11d, e). 2) The orebody and the fault offset in the sandstone-basement interface both result in relatively weak diffracted energy (hyperbolic trajectories labelled P-P and P-S in Fig. 11) emanating from the vicinity of the fault and the orebody. The amplitudes of the diffractions would be even weaker at the larger offsets in the case of a 3-D orebody (i.e. one of which has a limited strike-length) as the 2-D modelling, which assumes cylindrical symmetry, underestimates the amplitude reduction with offset (e.g. Sheriff and Geldart, 1995, p. 59). 3) An asymmetry in the amplitudes of the diffractions is observed. This is particularly true in the case of diffractions (both P-P and P-S) from the orebody due to the dip of the orebody, resulting in higher amplitudes observed in the downdip direction (compare Fig. 11b and c). 4) The P-to-S converted wave diffraction from the orebody is more extensive than the P-wave diffraction (see Fig. 11c). 5) The downdip P-to-S converted wave diffraction from the orebody is more prominent on the horizontal component (compare Fig. 11c and 11f). These simple modelling results indicate that the geometry of the seismic response of the McArthur River orebody (i.e. diffraction) is generally similar to the diffraction resulting from the hosting (or bounding) fault; however, the asymmetry in the amplitude response (higher amplitudes downdip) of the orebody has the potential to distinguish mineralized fault zones from barren faults. The predominance of orebody dip in determining the seismic response has been noted elsewhere (e.g. Bohlen et al., 23; Clarke and Eaton, 23). The characteristic diffraction response is best observed on the unmigrated section, suggesting that seismic profiles should be evaluated for potential orebody responses prior to migration (e.g. Adam et al., 1997; Milkereit et al., 1997). The larger response recorded on the horizontal component suggests that horizontal component geophones may provide enhanced detection of the steeply dipping ore zones. SEISMIC DATA The multielement seismic acquisition program (see Fig. 1, 2, and 3 for locations) was designed to address the objectives described in the Introduction section. All of the components listed in the Introduction are described below with the exception of the multicomponent 3-D VSP. Prior to presenting the various data sets, a brief discussion of seismic resolution is provided to allow valid comparison of the different seismic data sets, geophysical logs, and core measurements. The various seismic methods used in this study differ significantly in resolving power. The scale of geological features that can be seismically imaged is fundamentally determined by the frequency content of the seismic wavelet 372

11 D.J. White et al. V p = 48 m s -1-1 V s = 3394 m s ρ = 25 kg m -3 Sandstone 5 m V p = 58 m s -1-1 V s = 411 m s ρ = 3 kg m -3 Basement gneiss V = 54 m s -1 p -1 V s = 3818 m s ρ = 1 kg m -3 Ore zone Absorbing boundary 25 m Figure 1. The 2-D elastic model used for seismic simulation. The seismic properties of the orebody, sandstone layer, and basement are indicated. The model is defined by specifying elastic properties on a regular grid with grid nodes spaced at 2 m intervals, the time-step used for the finite-difference calculation was.2 ms, and the central frequency is 1 Hz. An explosive plane-wave source was used. A 2-D finite-difference elastic-wave algorithm was used for the simulation (Bohlen, 22). and by the seismic wavespeed (V p for compressional waves) of the subsurface. Conventionally, the vertical resolution is defined in terms of the minimum thickness of a thin layer that can be imaged, which for a seismic wave of wavelength λ is about λ /8 (Widess, 1973), where λ =V p /f c, and f c is the central frequency of the seismic wavelet. Estimates of achievable lateral resolution are defined by the Fresnel radius for unmigrated data (F = (λz/2) ½ ) where z is depth. (Fresnel radius is the radius of the area on a reflector from which reflected wave energy arriving at a detector contributes to the observed reflection by constructive interference.) For accurately migrated data (i.e. when the subsurface wavespeed V p is well known and aperture permits) the lateral resolution is λ. (Migration is a data manipulation process that repositions observed reflections from data recording co-ordinates to the locations in the subsurface from which the reflected (or diffracted) energy originates. For example, reflections from a dipping reflector are originally plotted (unmigrated) directly beneath the surface detector that records them, whereas the actual reflection point is located in the updip direction from this point. Upon migration, dipping reflections move updip and steepen, whereas diffractions are collapsed to a point. See Sheriff and Geldart (1995) for further details.) A summary of the frequency content of the different seismic methods used in this study and their resolution limits are provided in Table 4. High-frequency vertical seismic profiling The primary advantage of vertical seismic profiling (shown schematically in Fig. 12) is that observed reflections can be traced directly to the borehole where the geology is known. Seismic waves emanating from a surface source are reflected from geological interfaces and recorded by geophones located in the borehole. As the geophone locations approach the point where the borehole intersects the reflecting geological horizon, the reflection point on the horizon and geophone position converge. This geometry allows definite association of reflections with specific geological interfaces. Furthermore, traveltimes from the surface to known depths within the borehole provide accurate velocity information for conversion of the surface seismic data from traveltime to depth. Vertical seismic profiles were acquired using near- and far-offset surface sources located at distances of 27 m at N86 o W and 326 m at N124 o W from the MAC-218 borehole collar (see Fig. 3 for location), respectively. The data were acquired using a four-level, three-component geophone downhole seismic acquisition system to record energy generated by a mini-vibroseis system. The mini-vibroseis system is capable of producing a signal with 2 3 Hz bandwidth as compared to a maximum frequency of 17 Hz achieved in the high-resolution surface seismic survey (see section 2-D high-resolution seismic profiles ). The objectives in acquiring the VSPs were two-fold: to calibrate the surface seismic-reflection profiles, and to test the efficacy of high-frequency acquisition for resolving basin-fill stratigraphy. The acquisition parameters are provided in Table 5. The data sets for both VSPs were processed following a similar sequence (Table 6). Figure 13 shows the processed data for both VSPs, compared to the borehole geology and geophysical logs from Mwenifumbo et al. (27). Also shown is a comparison of V p versus depth as determined independently from the sonic log and near-offset VSPs first-arrival traveltimes. Figure 14 compares the log-based impedance, reflectivity, and synthetic seismograms, against 373

12 GSC Bulletin 588 Table 4. Estimates of vertical and lateral resolution for the various seismic methods employed in this study. A representative wavespeed of 4 m s -1 (appropriate for sandstone) has been used for the calculations and the Fresnel radius is determined for a depth of 5 m. Note that in the case of the sonic and laboratory measurements, vertical resolution refers to the length over which V p is measured (i.e. the distance between the source and receiver). Method Frequency band (Hz) Central frequency (f c ; Hz) Vertical resolution (m) Fresnel radius (m) Regional data High-resolution VSP Sonic 2.8x x1 4 ~.2 N/A.14 λ (m) Laboratory ~.5 N/A 4x1-3 N/A = not applicable Recording system Borehole with 4-level tool Vibrator Figure 12. Schematic diagram showing the vertical seismic profile (VSP) acquisition geometry. Table 5. High-frequency VSP acquisition parameters. Parameter Zero-offset VSP Offset VSP Source Mini-Vibroseis Mini-Vibroseis Sweep frequencies 2 3 Hz linear upsweep 2 2 Hz linear upsweep Sweeps per VP Source offset from collar 27 m 326 m Receiver spacing 2.5 m 5 m Depth range covered 6 46 m 6 46 m Number of recording levels Recording instrument Oyo seismograph Oyo seismograph Downhole tool 4-level Vibrometrics 4-level Vibrometrics the corridor stack and the VSP-CDP-transform depth image determined from the VSP data. (Corridor stack is the result of horizontally aligning upcoming waves recorded in each trace of the VSP and summing these traces to accentuate the primary reflections. The resultant reflectivity estimate can be compared directly to coincident surface seismic data. A VSP-CDP-transform is an image reconstruction technique that transforms the VSP data into the co-ordinate system of the surface seismic data (i.e. a vertical section). See Dillon and Thomson (1984) for further details.) The corridor stack procedure transforms the VSP data into a format that is more directly comparable to corresponding 1-D synthetic seismograms determined from borehole logs (e.g. Fig. 8). The far-offset VSP characterizes the reflectivity over distances of up to 15 m from the borehole, in contrast to the near-offset VSP that images only the immediate vicinity of the borehole. Referring to the VSP data in Figures 13 and 14, significant, but weak reflectivity is observed over the depth range of the basin-fill strata (MFb through MFd and RD), and reflectivity within the individual members of the Manitou Falls and Read formations is generally as strong as reflectivity associated with the boundaries between the members. This suggests that regional seismic mapping of the boundaries between the individual units will be difficult, although their internal reflectivity may provide a means of identifying the individual units. Reflections are generally not laterally continuous over large distances within the sandstone column (Fig. 13e). Notably, the reflectivity observed on the VSPs differs significantly from that determined from the sonic logs, and the reflectivity observed on the near- and far-offset VSPs is not in total agreement. The cause of these apparent discrepancies is discussed below. Despite the overall weak reflectivity within the Manitou Falls and Read formations, there are prominent reflections that can be related to specific horizons within the basin-fill strata. Vertical seismic profiling reflectivity to about 35 m depth (Fig. 14e, f) is relatively weak with the exception of a prominent reflection at about 21 m depth (A) and a band of reflectivity at about 3 m depth (B). The upper reflection (A in Fig. 14e, f) appears to register a significant diagenetic-hydrothermal feature, by comparison with borehole geophysical and clay mineralogy logs (Fig. 7), where an abrupt downward decrease in chlorite and dravite is compensated by an increase in illite; however, given the general insensitivity of V p to clay composition (Tosaya and Nur, 1982), this feature is interpreted 374

13 D.J. White et al. Table 6. High-frequency VSP processing sequence. Note that VSP-1 and VSP-2 refer to the near- and far-offset VSP data, respectively. Sequence Process 1 Reformat data from SEG2 2 Assign survey geometry 3 Sort data by wireline depth 4 Edit noisy and dead traces 5 Notch filter electrical noise (61.25 Hz, Hz, and Hz) 6 First break picks on vertical component data 7 Rotate 3-C data to maximize H1 horizontal component (using 2 ms window centred on first breaks) 8 Remove downgoing wave energy VSP-1: 11-point median filter for p-wave VSP-2: 9-point median filter for p- and s-wave; 7-point filter for tubewave 9 Remove residual downgoing wave energy using fk filter (VSP-1 only). 1 Bandpass filter: VSP-1: 4/7 17/2 Hz VSP-2: 2/4 15/17 Hz 11 Amplitude gain applied: 3-C automatic gain control to preserve the amplitude ratio between components 12 Corridor stack (VSP-1 only): -shift data to two-way traveltime using first breaks -apply top and bottom mute -stack as a (potentially fault-related) fracture zone based on the full waveform sonic logs (Fig. 7). The coincidental sharp clay mineral change is interpreted as fault-foreshortening of vertical clay zonation in and around the P2 mineralized zone (see also RL-88, Fig. 5 in Mwenifumbo et al. (27)), although it is recognized that similar abrupt changes in clay mineralogy often occur independent of fault offsets. A lower reflectivity band at about 3 m depth (B in Fig. 14e, f) appears to be associated with the strong silicification front, although there are different measures of the depth at which this gradational front actually begins. On the VSP velocity log (Fig. 13b) the front occurs at about 25 m depth as compared to about 3 m depth on the sonic log, whereas marked increases in density and resistivity are observed at m (Fig. 2 in Mwenifumbo et al., 27), and a strong silicification front was first noted in lithological logs by Yeo et al. (27) at m depth. Reflectivity in the VSP corridor stack and VSP depth image (reflections C to F in Fig. 14e, f) increases substantially in the vicinity of the unconformity (depth marked by UC), although the reflection from the unconformity itself is not pronounced. All of these reflections, with the exception of reflection F, lie above the unconformity based on correlation with the geological log (Fig. 14a). Reflections C and E are also observed on the log-generated synthetic seismograms and are confidently identified as zones of increased silicification within the sandstone column in the case of C and a low-density low-impedance zone near the unconformity at about 475 m depth (Fig. 14b) in the case of E. The last is a feature corroborated by multiparameter logs in some other drillholes (Mwenifumbo et al., 24). The absence of reflections D and F (which is deeper than the bottom of the log) on the log-based synthetic data precludes definitive identification of the source of these reflections. Based on earlier discussion, reflection D is likely a zone of low porosity due to increased silicification on a larger scale than sampled by the borehole. Reflection E, which lies within the basement at a depth of about 57 m, corresponds to a weak, but laterally continuous reflection on the VSP-2 depth image. It may be a reflection within the basement, or it could represent the base of the regolith. The regolith constitutes moderately altered semipelitic and pelitic gneiss units at the base of this borehole. The different scales of sampling between the sonic logs and the VSPs, result in incomplete agreement between the estimates of reflectivity from sonic-log-based synthetic seismograms and the VSP-1 corridor stack or VSP-2 depth image (Fig. 14). Whereas the geophysical logs provide a direct link between the geology and rock properties in the immediate vicinity of the borehole that determine reflectivity, the VSP results show which reflectors are extensive enough laterally to be observed on the surface seismic profiles. The importance of scale is further emphasized in Figure 13 where the VSP reflectivity for both the near- and far-offset VSPs are compared. Whereas the near-offset VSP samples the immediate vicinity of the borehole (i.e. within ~8 m as defined by the Fresnel radius; see Table 4), the far-offset VSP has reflection points that extend up to 15 m from the borehole. Thus, prominent reflections that are observed on both VSPs over a large depth range have enough lateral extent to be observed on the surface seismic profiles. Furthermore, significant differences are observed when comparing the interval velocities estimated from the VSP and sonic logs for MAC-218 (Fig. 13b). These differences would lead to a discrepancy of about 2 m (or 3%) in depth at the unconformity depth (~58 m) if the surface seismic data are alternatively converted to depth using the sonic or VSP velocities. The source of such traveltime or interval velocity (the average velocity determined over some corresponding depth range (or travel path)) discrepancies between sonic and VSP data are well documented (e.g. Stewart et al., 1984). They can be due to a combination of 1) intrinsic velocity dispersion that characterizes fluid-saturated porous media (e.g. Batzle et al., 21), 2) apparent velocity dispersion caused by short-period multiples (e.g. Schoenberger and Levin, 1974), or 3) scale-dependence of porosity. In the case of the first two causes, velocities generally increase with frequency, typically resulting in the average velocities being higher for the sonic log as compared to the VSP (Stewart et al., 1984). This is the case here where the mean velocity determined over the interval from 6 m to a given depth (not shown) from the sonic log is faster by 4 6% relative to the VSP velocity, and variations of up to 18% are observed in the interval velocities (Fig. 13b). Scale-dependent porosity refers to a situation where the bulk porosity is different for the scales sampled by the sonic log and VSP. This could occur if there are lateral variations in porosity, or large fracture porosity exists that is not sampled by the sonic log. Thus, given that the VSP frequency band is 375

14 GSC Bulletin 588 a) b) c) d) e) Sonic VSP V (km s ) p x1 6 Z (kg s m ) First breaks Sandstone column Basement Depth (m) Depth (m) Formations MFd MFc MFb RD Basement Figure 13. Comparison of geology, geophysical logs, and near- and far-offset VSP data for MAC-218. a)geological log (from Cameco Corporation, unpub. data, 2: RD= Read Formation (former MFa); MFb, MFc, MFd indicate the different members of Manitou Falls Formation; dark grey below RD is basal conglomerate part of RD resting on basement). b) P-wave (sonic) velocity log (Mwenifumbo et al., 2) and interval velocities determined from the VSP first-arrival traveltimes. c) Log-based calculated acoustic impedance (Mwenifumbo et al., 2). d) Near-offset and e) far-offset processed VSP data. Note that each trace in the VSP data has been shifted in time so that the first arrival occurs at twice the actual measured first-arrival traveltime (dashed line labelled first breaks ). This process aligns reflections from horizontal reflectors in the VSP data so that each reflection occurs at a constant two-way traveltime (vertical trajectories as plotted in Fig. 13d, e). This simple transformation is accurate for the near-offset VSP, but only approximate for the far-offset VSP. A proper transformation of the far-offset VSP data is shown in Figure 14. The dashed lines indicate the depths at which some of the more prominent reflections occur. 376

15 Basement RD MFb MFc Depth (m) x16 Z (kg s-1 m-2) -4 4 x1-3 Reflectivity 5 4 Synthetic seismogram C E Depth from sonic Vp F D E UC C B Distance from borehole (m) Depth from sonic Vp f) Figure 14. Comparison of borehole logs and VSP data for MAC-218. a) Geological log (from Cameco Corporation, unpub. data, 2), b) acoustic impedance calculated from Vp and density logs, c) reflectivity calculated from impedance, d) synthetic seismogram for 2 3 Hz Klauder wavelet, e) near-offset VSP corridor stack, and f) VSP-CDP-transform depth section. Note that the VSP images in Figures 14e and 14f are converted to depth using the velocities from the sonic logs to make valid comparison with the sonic log synthetic seismograms. The labelled reflections are discussed in the text A 1 1 e) 1 d) 1 c) 1 Depth (m) MFd b) Depth (m) a) Depth (m) Depth (m) Formations 15 F UC D Contents Author Index D.J. White et al. 377

16 GSC Bulletin 588 closer to the surface seismic band (see Table 4), where available, VSP data should be used to adjust the sonic logs for correct comparison to the seismic data. 2-D high-resolution seismic profiles Two high-resolution 2-D profiles (lines 12 and 14; see Fig. 3 for location) were acquired across the northeastern end of the P2 fault zone, transecting the McArthur River mining camp. Detailed acquisition parameters, data processing sequence, and a detailed tectonostratigraphic interpretation of migrated versions of these profiles can be found in Györfi et al. (27). Unmigrated images are shown in Figure 15 (see colour folio, p. 64). Here, the authors focus on the imaging capability of these high-resolution profiles in terms of the local geology and within the environment of an operational mine site. Signal-to-noise analysis Acquisition of seismic data in the vicinity of an active mine operation presents some challenges due to the relatively high ambient acoustic noise levels within the seismic frequency band. The synthetic modelling of the section Reflectivity of the orebody, along with data-based signal-to-noise ratio (S/N) estimates, can be used to gain insight regarding spatial variations in image quality (i.e. do they represent geological variations or recording conditions?), and the likelihood of observing a direct response from the orebody. Figure 16 (see colour folio, p. 65) shows the S/N for all stations of line 12, and Figure 17 shows the mean S/N at each recording (or common depth point (CDP)) station along both lines 12 and 14. In Figure 16, valid comparison of S/N can be made for similar shot-to-receiver offsets (i.e. trajectories running parallel to the bright (red-to-yellow) diagonal zone). The signal-to-noise ratio decreases with offset as would be expected due to the 1/distance decay of amplitude for a spherical wavefront. The signal-to-noise ratio is highly variable along the profiles with mean values ranging from 4 db to 14 db (Fig. 17; recall that each 3 db represents reduction by a factor of 2) with most values in the 6 9 db range, and S/N is generally higher along line 12 than line 14, generally consistent with the difference in clarity between the corresponding images. Low S/N is observed in the immediate vicinity of the mining camp (near receiver or CDP stations 3 5 along line 12), with the highest values occurring well outside of the camp (southeast end of line 12). The nominal data fold for profiles 12 and 14 is 5 6, and thus the S/N enhancement that may be achieved by stacking is optimistically a factor of 7 8 (theoretically N 1/2 where N is the number of traces being stacked) or about 9 db. True-amplitude plots of example shot gathers are shown in Figure 18 (see colour folio, p. 66). Where S/N levels are high (Fig. 18a), the observed reflection from the unconformity zone (S2 in Fig. 18a; amplitude about 24 db down) exceeds the background noise (3 34 db down) by about 6 db in the region outside of the strong ground roll signal. (Ground roll refers to the surface wave that propagates horizontally along the free surface of the Earth. It often constitutes a large amplitude coherent signal that often obscures reflected arrivals.) The unconformity zone reflection (S2) is clearly seen in the unstacked data. Where S/N is reduced (Fig. 18b) the noise floor (2 24 db) exceeds the unconformity zone reflection strength by up to 4 db. Thus, although the reflector is not observed in the shot gather, it should be observable after stacking due to the expected nominal 9 db enhancement. Considering that the theoretical strength of the orebody response (see data simulation in Fig. 11) is about 6 db less than the unconformity zone reflection (say ~3 db down in terms of the amplitudes of Fig. 18), observation of a direct response from the orebody in the unstacked data would be marginal at best. After stacking, a direct response should NW Signal/noise (db) NW Signal/noise (db) a) b) Line 12: S/N (db) vs. cdp station Common depth point station Line 14: S/N (db) vs. cdp station Common depth point station SE SE Figure 17. Mean S/N levels versus receiver station for a) line 12 and b) line 14. Signal-to-noise ratio levels determined in a 3 ms window prior to and following the first arrival, respectively. The dashed lines delineate a central region where mean S/N levels can be validly compared. Outside of this region, the variability of the source-receiver offset distribution affects the noise estimate. 378

17 D.J. White et al. be observable in high S/N areas (e.g. conditions away from the mining camp, southeast end of line 12), but would be unlikely in low S/N areas (i.e. the mining camp). The importance of recording data at source-receiver offsets that are large relative to the target depth is appreciated by noting that along much of the length of the high-resolution profiles (particularly on line 14, but also on line 12) the S2 reflection observed in the stack comes primarily from large offset traces. Inspection of Figure 18 depicts how the S2 reflection is only observable once it emerges from the part of the gather that is dominated by ground roll. This is the case for the majority of shot gathers for both profiles 12 and 14. In optimal circumstances (e.g. along southeast line 12 where S/N is high, and ground roll is reduced) the S2 reflection can a) be followed from the inside traces to far offsets. This also emphasizes the importance of applying ground roll suppression techniques. Images from lines 12 and 14 Unmigrated versions of the 2-D high-resolution profiles with a basic interpretation superposed are shown in Figure 15. Lateral variation in the overall clarity of the seismic images is largely due to variable recording conditions, both ambient background noise and severity of source-generated ground roll as discussed in the previous section. This is particularly apparent along line 12 where the image is clear along the southeast half of the line that is characterized by high S/N b) c) S1 S2 S2 d) NW F Projection 258 m F S2 SE S2 Rotated sandstone beds S2 e).2.4?p2 48 Depth (m) 96 Figure 19. a), b), c) Expanded view of selected data windows from Figure 15, to demonstrate the variability in the unconformity reflection (S2). Also shown are d) interpreted rotation of sandstone beds due to multiple reverse faulting (Györfi et al., 27) and e) a locally bright reflection near the P2 mineralized zone. 379

18 GSC Bulletin 588 levels and less severe ground roll (see Fig. 17, 18). Figure 19 provides an expanded view of example features of the seismic data. Here, the focus is on simple interpretation of the first-order features of the seismic data; the reader is referred to Györfi et al. (27) for more comprehensive geological interpretations. A prominent feature observed on both profiles (Fig. 15) is a laterally semicontinuous reflection (S2) at about.2.3 s two-way traveltime (~4 6 m depth) that divides the sections vertically into two domains of distinct reflectivity. The upper domain has generally subhorizontal or gently undulating reflectivity, whereas the lower domain has generally shallowly southeast-dipping or more erratic reflectivity. Based on the truncation of underlying dipping reflections (B) at S2, the S2 reflection is interpreted as originating from the unconformity zone. This interpretation is corroborated by depths to basement from limited borehole data (locations shown in Fig. 15), the predicted (see Seismic reflectivity of the basin rocks ) and the observed (see Fig. 14) prominent reflectivity in the vicinity of the unconformity. In the one location where reflectivity is calibrated by VSP data (borehole MAC-218; Fig. 14 and Comparison of line 14 and MAC-218 VSP images section), the most prominent reflectivity begins at about 4 m above the unconformity as logged in drill core. Thus, to be prudent, basement in Figure 15 has been interpreted as the base of the prominent reflector S2 and the term unconformity zone is used. This zone will variably consist of silicified basal sandstone and/or conglomerate, regolith, and fractured basement. Reflection S2 shows considerable vertical relief (up to.5 s or ~12 m) along both profiles, forming broad open undulations (e.g. CDPs 65 8 along line 12) and abrupt offsets (F1 in Fig. 15a). The latter are most reasonably interpreted as fault offsets and can be followed in some cases (e.g. F1 in Fig. 15a, 19b) as relatively low-angle structures into the underlying basement. The P2 fault (P2 in Fig. 15a, b), which is associated with the McArthur River orebody at this location, can be seen as a vertical offset along the interpreted unconformity, and can be followed well into the basement (at least on line 12) where it occurs as a prominent southeast-dipping reflection. Györfi et al. (27) interpreted the P2 fault as a reactivated basement shear zone, analogous with interpretations by Tourigny et al. (22, 27) of the Sue trend. The undulations and abrupt offsets observed in the seismic images are consistent with either the existence of paleotopography (i.e. presedimentation relief) and/or syn- to postsedimentation faults. Reflection S2 also shows considerable lateral variability in character as highlighted in Figure 19. It ranges from being a strong laterally continuous reflector (Fig. 19a), strong, but laterally segmented (i.e. faulted; Fig. 19b), to weakly reflective (Fig. 19c). This variability is certainly due to the geological variations in the vicinity of the unconformity (see Seismic reflectivity of the basin rocks for discussion). In the case of Figure 19a, the pronounced lateral coherency of the reflection suggests that it is likely due to a continuous stratigraphic unit at the base of the sandstone column (possibly silicified sandstone or tabular conglomerate) as unconformities are not commonly this pronounced (e.g. Dietrich, 1999; Ouassa and Forsyth, 22). Where S2 is weakly reflective, it may indicate that the regolith represents a gradual transition to unweathered basement rocks and alteration is limited. Reflectivity above S2 within the Athabasca Basin fill is generally highly variable both vertically and laterally, with many of the reflections having lateral coherence of less than 1 2 m. An exception to this is a shallow, laterally continuous reflection (S1, line 12 in Fig. 15) at about.5 s two-way traveltime (or ~12 m depth) that has relatively little relief (<.2 s or 5 m) and minimal evidence for significant fault disruption over its about 1.5 km lateral extent. It correlates well with a shallow high-density horizon observed in MAC-259. Short subhorizontal reflections within the stratigraphic column likely represent lateral variations in the physical properties of the strata in many cases (e.g. silicification, porosity, etc.; see Mwenifumbo et al. (2) and discussion in Seismic reflectivity of the basin rocks ) or perhaps facies changes. In some instances, discontinuities in sandstone reflections (e.g. A in Fig. 15a) suggest angular unconformity or broad fluvial channel scours and fills within the stratigraphic column. Györfi et al. (27) interpreted internal discontinuities located at different stratigraphic levels within the sandstone as either offsets associated with multiple fault splays (e.g. Fig. 19d), or local unconformities resulting from tectonic control on sedimentation. The internal sandstone discontinuities and the complex image of the P2 fault are consistent with interpretations by Bernier et al. (21) of variable syndepositional faulting related to the P2 fault, and the anastamosing braided channel architecture of these kinds of strata (Bernier et al., 21). A short, relatively strong reflection is observed at s two-way traveltime (or ~5 55 m) near CDP 4 of line 12 (Fig. 15, 19e) that corresponds to the mineralization trend that lies along strike from the McArthur River ore zones (see Fig. 3 for location). This locally bright reflection is tentatively interpreted as an ore-related response, although it is uncertain what particular aspect of the mineralized zone is responsible for producing this response. Based on its depth, it appears to be located in the footwall of the P2 fault. To better estimate the true dips of the basement reflectors that are imaged, the unmigrated image from Figure 15a has been projected onto a vertical plane that has a strike that is subparallel to the estimated regional geological strike (315 ). Basement reflections (B in Fig. 2) dip at projected (but unmigrated) angles of less than 27 SE, which will steepen to less than 31 SE when data are migrated. These dips are generally shallower than the regional 4 45 dips observed in the basement. Farther northwest, basement is less reflective, indicating either more structural complexity, a change in structural attitude to steeper dips, or a change in basement lithology. The maximum dip angle that can usually be resolved in seismic-reflection surveys of this type is about 6. Comparison of line 14 and MAC-218 VSP images The ultimate objective of acquiring the VSP data was to allow direct correlation of the surface seismic images to the geology from borehole MAC-218. The VSP corridor stack 38

19 D.J. White et al. NW 1 Station number A S1 SE.2 S P2 F1 B Depth (m) Figure 2. Unmigrated data has been projected onto a section at 315, in an attempt to better represent the true dips within the basement. Labels are referred to in the text, and are the same as those shown in Figure 15a. The grey shading in the upper part of the figure indicates the interpreted basin-fill sediments. The dashed lines represent interpreted faults. The labels are defined within the text. and VSP-CDP-transform image are compared with the seismic image for line 14 in Figure 21. The VSP data have been low-pass filtered (cf. Fig. 14) to better match the frequency content of the surface seismic data, although the surface data are still characterized by a lower dominant frequency (~5 Hz). Inspection of the line 14 high-resolution seismic-reflection image at the projected position of borehole MAC-218 reveals a very complex structure making rigorous correlation difficult; however, the transition from the sandstone strata to the crystalline basement (including the unconformity) is clear on the filtered VSP image and corridor stack and correlates very well with the laterally continuous unconformity zone reflection interpreted on the line 14 section (Fig. 15). The top of this transition zone (reflection D on the corridor stack and VSP image) correlates with the top of the unconformity reflection that was interpreted as a silicified zone at the base of the sandstone column (see Images from lines 12 and 14'). The reflections below this are a superposition of individual reflections from the geological features (low-density zone E, unconformity UC) identified in Figure 14, which is a consequence of the relatively long wavelengths of the surface seismic data (~1 m) compared to the vertical extent of these features. The lower frequencies also accentuate the contrast in reflectivity of the sandstone column relative to the sandstone-basement transition. The prominent reflection on the line 14 profile located above (at ~.16 s) the unconformity reflection likely corresponds to the large silicification-related porosity decrease observed at 3 m depth in Figure 14. As noted earlier, reflection E lies within the basement and may be a reflection from the base of the regolith. 3-D high-resolution seismic survey A high-resolution 3-D survey was conducted to obtain constraint on the true geometry of structures imaged by the 2-D high-resolution profiles in the immediate vicinity of the mining camp. Initial design considerations for the limited 3-D survey were described in Hajnal et al. (2). The actual survey geometry achieved and the resultant fold map is shown in Figure 22 (see colour folio, p. 67). The 3-D coverage was obtained by recording shots from the 2-D lines (12 and 14; Fig. 3) on a series of cross-receiver lines (1, 4, and 6), as well as recording a series of short auxiliary shot lines (1, 3, 5, 7, 9, 11, 13, and 15) on these receiver lines and by receivers on the 2-D lines (12 and 14). The acquisition parameters (similar to the 2-D high-resolution parameters) are provided in Table 7. Combining the standard cable geophone channels and the VectorSeis geophones (deployed on lines 4 and 6), a total of almost 16 vertical component recording channels were available for the 3-D survey. 381

20 GSC Bulletin 588 Station number Distance from borehole (m) Station number A B C UC D F Corridor stack C B A E D E UC F VSP D UC F Figure 21. Comparison of line 14 surface seismic data, VSP corridor stack, and offset-vsp depth image. The depth of the various horizons (A-F) identified by comparison with borehole geology (from Fig. 14) are shown. The surface trace of the VSP depth image plane is oriented at 225 making it approximately perpendicular to the line 14 surface seismic profile. The right and left sides of the VSP image are located about 275 m and 125 m from line 14, respectively. Note that the corridor stack and VSP image are both presented in time and have been filtered with a passband of 5 Hz to 9 Hz to better match the frequency content of the surface seismic data. The corridor stack has been corrected to the same datum (-25 ms bulk shift) as the surface data, allowing direct comparison. The VSP image has been positioned to match the corridor stack reflections. Table 7. Two-dimensional and three-dimensional seismic acquisition parameters. Parameter 2-D high-resolution 2-D regional 3-D high-resolution Recording instrument Source type IO-system 2 24-bit telemetry, with noise burst edit and diversity stack IVI Y kg Vibroseis buggies, 47 7 lbs peak force per unit IO-system 2 24-bit telemetry, with noise burst edit and diversity stack IVI Y kg Vibroseis buggies, 47 7 lbs peak force per unit IO-system 2 24-bit telemetry, with noise burst edit and diversity stack, and Vectorseis remote seismic recorders IVI Y kg Vibroseis buggies, 47 7 lbs peak force per unit Number of vibes Vibration point 2 m 5 m, 25 m for 3 km at line 2 m (VP) interval ends Sweep frequencies 3 17 Hz nonlinear (3 db/octave) upsweep 1 84 Hz linear upsweep 3 17 Hz nonlinear (3 db/octave) upsweep Number of 4 6 (4 vibes) or 1 (3 vibes) 4 sweeps per VP Sweep length 12 s 28 s 12 s Record length 6 s 18 s 6 s (correlated) Geophone group interval 5 m 25 m 5 m vertical component groups 4.2 m VectorSeis 3-component Geophones per group 6 over 5 m 6 over 25 m 6 over 5 m, vertical component groups; 1 at station, VectorSeis 3-component Vertical 1 Hz 1 Hz 1 Hz geophone type Sample interval 4 ms 1 ms 1 ms Number of recording channels 96 vertical component 96 vertical component 96 vertical component 6 VectorSeis 3-component digital phones 382

21 D.J. White et al. Data from the 3-D survey (White et al., 21) have been processed to obtain a 3-D data cube using the processing sequence as indicated in Table 8. The results must be considered in view of the degradational effects of uneven offset distribution and azimuthal coverage within the CDP bins introduced by the limited 3-D acquisition geometry (e.g. Stone, 1994). Specifically, the uneven offset distribution compromises the cancellation of nonreflection energy during the normal moveout and stacking procedure. Data from cross-strike and along-strike vertical slices through the migrated data cube are shown in Figure 23 (see Fig. 3 or 22 for the location of these depth slices). There is significant energy within the data cube at about s two-way traveltime. This depth correlates well with the depth of the interpreted unconformity zone reflection where the 3-D sections tie with 2-D lines 12 and 14 (compare with Fig. 15). The offset observed along the interpreted unconformity zone reflection on the cross-strike section is consistent with that known for the P2 fault in both sense and magnitude. On the along-strike section the unconformity zone reflections deepen toward line 14, and the known locations of the ore zones correspond to zones of disruption in the interpreted unconformity zone reflection. Thus, accounting for the limitations of the 3-D acquisition geometry as implemented, it appears that 3-D seismic imaging is viable and shows potential for providing subsurface ore zone delineation. Crustal structure from the regional seismic surveys The two deep-sounding seismic-reflection profiles (lines A and B, Fig. 2) transect the P2 Main and P2 North mineralized zones. Detailed acquisition parameters, data processing sequence, and a detailed tectonic interpretation of these profiles are in Hajnal et al. (27). Here, the present authors consider the uppermost part of the section ( 2 s) of line B-B', (Fig. 24, see colour folio, p. 68). As in the case of the high-resolution lines, a laterally semicontinuous subhorizontal reflection (S2) in the shallowest part of the section (at ~.2.3 s) is interpreted as the unconformity zone reflection, as it vertically separates a weakly reflective zone above it (Athabasca Group) from the more highly reflective underlying basement. These reflections are clearly observed along the northwestern Table 8. Three-dimensional processing sequence. Sequence Process 1 Assign survey geometry, 15 x 15 m bins 2 Kill noisy traces 3 Spectral balancing (3 14 Hz) 4 5 ms automatic gain control (AGC) 5 First break picks 6 Top mute; 4 ms after first break Refraction statics (5 m datum; 45 m s -1 replacement 7 velocity) 8 CDP sort 9 NMO ( s: 45 m s -1,.2 s: 47 m s -1,.3 s: 5 m s -1,.5 s: 55 m s -1 ); no stretch mute 1 Bandpass filter (3 8 Hz) 11 Offset-dependent top mute 12 Stack (1 15 m offsets) 13 Fxy deconvolution 14 Migration half of the profile, are less clear within the central and southern region of the line due to interpreted structural complexity, yet the overall difference in reflectivity between Athabasca Group and basement remains strong throughout. This image of the unconformity zone clearly indicates a northwesterly increase in thickness of the Athabasca Group sandstone units from 4 m to 6 m. Significant topography is observed along the S2 reflector, including several abrupt offsets (F1, F2, F3, F4, P2) that have associated diffracted energy (unmigrated data) in some cases. Generally the sense of displacement along these interpreted faults (within the plane of the section) is southeast over northwest, consistent with the P2 fault; however, some faults (F1) have an apparent normal sense of displacement (down to the southeast). a) b) NW SW 5 m P2 fault Bin number Ore zones 5 m SE Bin number NE Figure 23. a) In-line, and b) crossline sections through the migrated 3-D data cube from the pseudo-3-d seismic survey. Locations of the data sections (3-D-2 and 3-D-1, respectively) are indicated in Figures 22 and 3. The approximate vertical exaggeration in these sections is.62:1. The horizontal arrows identify the interpreted unconformity reflection and the interpreted trajectory of the P2 fault is indicated by the arrow pointing updip in Figure 23a. 383

22 GSC Bulletin 588 A prominent southeast-dipping, approximately 25 m thick band of reflectivity (P2 in Fig. 24) extends from about.25 s to 2.3 s two-way traveltime. This zone is interpreted as an image of a southeast-dipping array of shear and/or fault zones, the uppermost fault of which (interpreted as the P2 fault) projects to the near-surface location of the McArthur River ore zone. This array of faults transects gently curved patterns that are interpreted by Hajnal et al. (27) as parts of basement fold structures, suggesting that this fault array is a relatively young tectonic feature, likely reactivated from a previous structural zone of ductile faults that were part of the ductile folding, analogous to the structural sequence detailed by Tourigny et al. (22, 27) in 3-D outcrop of the Sue C pit. In addition to the seismic image obtained along line B-B', tomographic inversion of the first-break (refracted) traveltimes from this line identifies a distinct high-velocity zone within the sandstone column in the hanging wall above and adjacent to the P2 fault zone (Fig. 24). This velocity anomaly is interpreted as representing a broad zone of hydrothermal silification of the sandstone as recognized in nearby lithological and geophysical borehole logs (McGill et al., 1993; Mwenifumbo et al., 21, 27). The regional seismic image from line B-B' is compared to the parallel high-resolution image from line 12 (Fig. 25, see colour folio, p. 69) that details the central part of line B-B'. The broad-scale features of this section are apparent in both images (i.e. the unconformity zone reflection and the southeast-dipping basement fabric that is here related to the P2 and related faults), although they are obviously much clearer in the high-resolution image. Although the depth variations in the unconformity zone reflection are traceable on the regional image, fault offsets are less clear, and the regional image provides very little detail within the overlying Athabasca Group sandstone units. It should be noted however, that this section of line B-B' is perhaps the lowest quality segment along the line (see Fig. 24). Furthermore, the regional data were processed with a view to providing a good crustal image from to 16 s. The quality of the shallow portion of the line B-B' image could likely be improved with processing focused on the upper 2 s. FINAL ASSESSMENT AND CONCLUSIONS To conclude, the authors summarize how well the objectives of the seismic program were met, with suggestions for followup work. Objective 1: define the subsurface stratigraphy of the sedimentary rocks within the basin The seismic data have shown limited use for directly distinguishing individual stratigraphic units of the Athabasca Group sandstone units (e.g. Manitou Falls Formation MFb, MFc, Mfd; and Read Formation RD); however, within the sandstone column, the high-resolution images do outline brittle deformation zones (and accompanying bed rotation), constrain syn- versus postsedimentary deformation, suggest stratigraphic disconformity or lateral facies changes, and in VSP logs clearly indicate interfaces of diagenetic alteration. The regional seismic images provide little information on the sandstone column, but could possibly be improved by focused processing of the shallow data. Objective 2: provide a detailed image of the basement unconformity The basement unconformity zone is clearly mapped by the regional profiles, and in excellent detail by the highresolution seismic profiles. The reflections that define the unconformity zone are a superposition of individual reflections from geological features occuring at the unconformity that may include basal silicified sandstone and conglomerate, fault-damaged zones, the unconformity proper, and the underlying regolith. Observed lateral variations in the reflection character of the unconformity zone are related to associated changes in the unconformity geology. The depth to basement, topographic relief, fault offsets, and sense of displacement on basement faults are clearly imaged in most places. Objective 3: characterize the basement unconformity using seismic attributes Variations in the seismic character of the unconformity zone reflection are unmistakable, but the geological significance of these variations is weakly constrained due to the paucity of drillholes that are located within 15 m of the seismic line, penetrate the basement, and sample the geology away from the anomalous mineralized zones. A detailed understanding of the seismic data within the Athabasca Group and unconformity zone will grow over time if such drillholes are drilled and logged with multiparameter tools, increasing the predictive value of the seismic images as a mapping and exploration tool. Objective 4: locate and image faults that have been instrumental in ore deposition Numerous fault zones in addition to the P2 structure are clearly identified on the seismic sections where abrupt vertical offsets in the reflection from the unconformity zone are observed. In addition, indicators on the seismic images uniquely identify the P2 fault zone, and may be diagnostic of mineralized fault zones. Perhaps the most significant indicator of the distinct nature of the P2 fault comes from the regional seismic images, where a much broader interpreted basement shear zone extends deep into the basement. This is reminiscent of the deep-plumbing fault systems that are associated with other types of hydrothermal mineral deposits (e.g. Lydon, 1995; Robert, 1995; Goleby et al., 1997; Drummond et al., 1998; Cassidy et al., 2). In addition, a high-velocity zone and locally bright reflection associated with the P2 fault may also be favourable indicators. 384

23 D.J. White et al. Objective 5: determine the seismic signature of a known ore deposit The authors saw no compelling evidence of the signature of the McArthur River orebody on the seismic data. The 3-D data do suggest that the unconformity is highly disrupted in the vicinity of the known ore zones. Synthetic modelling and signal-to-noise analysis of the high-resolution seismic data suggest that under ideal recording circumstances it may be possible to see a direct response from the orebody; however neither the regional nor detailed seismic lines could be shot closer than 3 m to a known ore zone, and attempts to obtain a direct image of the orebody through 3-D acquisition were hampered by high noise levels in the immediate vicinity of the mining camp. Objective 6: define the regional basement structure underlying the basin, including faults. Regional data calibrate a northwesterly increase in depth to basement from 4 m to 6 m over the length of the 25 km regional profile, identify basement faults, image a prominent interpreted basement shear zone associated with the P2 fault, and image a regional bright reflector that is interpreted as a regional sill that may have genetic affiliation with the basin development and subsequent mineralization (not discussed here, but see Hajnal et al. (27)). Recommendations for followup work 1) The most obvious gap in the existing work is the inability to define internal stratigraphy in the Athabasca Group. On the other hand, very strong imaging of the unconformity zone is the most important single exploration criterion, and clearly a success. Also very significant are the VSP results that identify the silicification front as a potentially strong reflector. These are key geological variations that may be relevant to identifying prospective ore zones, and should be carefully analyzed in the high-resolution transects and 3-D data. Further geophysical logging and VSP acquisition are required in regions where these seismic features are clear (e.g. southeast line 12, or where the unconformity zone reflection is well defined along line 14), and in new drillholes that are within 1 m of any of the now acquired seismic data. 2) Based on visual correlation with existing multiparameter geophysical and stratigraphic logs, it appears unlikely that a systematic seismic stratigraphic characterization can be developed within the Manitou Falls Formation; however, a more rigorous multivariate statistical characterization of the logs is required to fully assess the relationships between porosity, lithology, and seismic impedance characteristics of the individual stratigraphic units. It may be possible to show that the seismic fabric is parallel to primary Manitou Falls Formation stratigraphy, providing a better substantiated basis for extrapolating stratigraphic picks laterally from drillholes. Although sonic logging and rock-property measurements provide insight into reflectivity of the subsurface, scaling effects limit their accuracy, whereas VSPs provide a closer representation of reflectivity in relation to the surface seismic profiles. 3) Further work is warranted in assessing the seismic effects of clay content within the sedimentary column. Variations in fractional clay content have relatively minor effects on the seismic response because of the low values at McArthur River, but fractional clay content may be more significant when considering classic unconformity-type deposits (e.g. Cigar Lake) where clay fractions are much higher. 4) Laboratory measurements on cores from the sedimentary column demonstrated a strong dependance of acoustic impedance on porosity. Further work should be done in estimating porosity using the resistivity logs and comparing these logs to the V p logs to assess the relationship between porosity and reflectivity within the sedimentary column. Also, in that silicification and/or desilicification is proposed as the primary control on porosity, a suitable way of logging degree of silicification should be developed for quantitative comparison with the geophysical logs. This would help resolve existing ambiguity between stratigraphic and hydrothermal or diagenetic controls on reflectivity. 5) Given the inability to acquire a seismic line directly over the orebody in order to determine the seismic response of the ore zone, it would be useful to obtain a VSP in a borehole directly above the orebody. This would be feasible, as the source point can be offset several hundred metres from the borehole location. 6) The orebody modelling in this study was two-dimensional and based on the geometry of only one of the zones. Further modelling with accurate geometric geological and rock-property data is required to account for the 3-D variability in the characteristics of the ore zones. 7) Further modelling of the seismic response of the unconformity zone may be helpful to determine if waveform variations in the seismic data may be diagnostic of specific geological variations. 8) The success of the limited 3-D survey in this study suggests that a true 3-D survey would be highly beneficial in orebody and structural delineation. Considering the ambient recording conditions and the desire to maximize the value of such information, the best time to acquire such data would be late in the delineation drilling stage of developing new deposits. ACKNOWLEDGMENTS The authors thank D. Snyder for reviewing an earlier version of this manuscript. The Cameco Corporation staff at McArthur River are recognized for their co-operation during the seismic surveys. Special thanks also to Vlad Sopuck, Scott Mchardy, Garth Drever, Darcy Hirsekorn, and Garnet Wood. 385

24 GSC Bulletin 588 REFERENCES Adam, E., Arnold, G., Beaudry, C., Matthews, L., Milkereit, B., Perron, G., and Pineault, R. 1997: Seismic exploration for VMS deposits, Matagami, Quebec; in Geophysics and Geochemistry at the Millennium, Proceedings of Exploration 97: Fourth Decennial International Conference on Mineral Exploration, (ed.) A.G. Gubins; Prospectors and Developers Association of Canada, Toronto, Ontario, p Batzle, M., Hofmann, R., Han, D.-H., and Castagna, J. 21: Fluids and frequency dependent seismic velocity of rocks; The Leading Edge, v. 2, p Bernier, S., Jefferson, C.W., and Drever, G.L. 21: Aspects of the stratigraphy of the Manitou Falls Formation, Athabasca Basin, in the vicinity of the McArthur River uranium deposit, Saskatchewan; preliminary observations; in Summary of Investigations 21, Volume 2, Part b: EXTECH IV Athabasca Uranium Multidisciplinary Study; Saskatchewan Geological Survey, Saskatchewan Energy and Mines, Miscellaneous Report b, p (CD-ROM). Bohlen, T. 22: Parallel 3-D viscoelastic finite difference seismic modelling; Computers and Geosciences, v. 28, p Bohlen, T., Müller, C., and Milkereit, B. 23: Elastic wave scattering from massive sulfide orebodies: on the role of composition and shape; in Hardrock Seismic Exploration, (ed.) B. Milkereit, D. Eaton, and M. Salisbury; Society of Exploration Geophysicists, Tulsa, Oklahoma, p Card, C.D., Pan ( a, D., Stern, R.A., and Rayner, N. 27a: New insights into the geological history of the basement rocks to the southwestern Athabasca Basin, Saskatchewan and Alberta; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Card, C.D., Pan ( a, D., Portella, P., Thomas, D.J., and Annesley, I.R. 27b: Basement rocks to the Athabasca Basin, Saskatchewan and Alberta; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Cassidy, K.F., Goleby, B.R., and Drummond, B.J. 2: Imaging crustal-scale fluid pathways by seismic reflection; implications for Archean orogenic gold mineralizing systems; in Geological Society of America, 2 Annual Meeting, Abstracts with Programs, v. 32, p Clarke, G.J. and Eaton, D.W. 23: Influence of morphology and surface roughness on the seismic response of massive sulfides, based on elastic-wave Kirchoff Modeling; in Hardrock Seismic Exploration, (ed.) B. Milkereit, D. Eaton, and M. Salisbury; Society of Exploration Geophysicists, Tulsa, Oklahoma, p Dietrich, J.R. 1999: Seismic stratigraphy and structure of the Lower Paleozoic, Central Alberta LITHOPROBE Transect; Bulletin of Canadian Petrolum Geology, v. 47, p Dillon, P.B. and Thomson, R.C. 1984: Offset source VSP surveys and their image reconstruction; Geophysical Prospecting, v. 32, p Drummond, B.R., Goleby, B.R., Goncharov, A.G., Wyborn, L.A.I., Collins, C.D.N., and MacReady, T. 1998: Crustal-scale structures in the Proterozoic Mount Isa Inlier of north Australia: their seismic response and influence on mineralisation; Tectonophysics, v. 288, p Eberhart-Phillips, D., Han, D.-H., and Zoback, M.D. 1989: Empirical relationships among seismic velocity, effective pressure, porosity, and clay content in sandstone; Geophysics, v. 54, p Fouques, J.P., Fowler, M., Knipping, H.D., and Schimman, K. 1986: The Cigar Lake uranium deposit: discovery and general characteristics; in Uranium Deposits of Canada, (ed.) E.L. Evans; Canadian Institute of Mining and Metallurgy, Special Publication 33, p Goleby, B.R., Drummond, B.J., Owen, A., Yeates, A.N., Jackson, J., Swager, C., and Upton, P. 1997: Structurally controlled mineralization in Australia how seismic profiling helps find minerals: recent case histories; in Geophysics and Geochemistry at the Millennium, Proceedings of Exploration 97: Fourth Decennial International Conference on Mineral Exploration, (ed.) A.G. Gubins; Prospectors and Developers Association of Canada, Toronto, Ontario, p Györfi, I., Hajnal, Z., White, D.J., Takács, E., Reilkoff, B., Annesley, I.R., Powell, B., and Koch, R. 27: High-resolution seismic survey from the McArthur River region: contributions to mapping the complex P2 uranium ore zone, Athabasca Basin, Saskatchewan; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Hajnal, Z. and Reilkoff, B.R. 198: Reflection data processing in the Athabasca basin; in Expanded Abstracts, 5th Annual International Meeting of the Society of Exploration Geophysicists, p Hajnal, Z., Annesly, I.R., White, D., Matthews, R.B., Sopuck, V., Koch, R., Leppin, M., and Ahuja, S. 1997: Sedimentary-hosted mineral deposits: a high-resolution seismic survey in the Athabasca Basin; in Geophysics and Geochemistry at the Millennium, Proceedings of Exploration 97: Fourth Decennial International Conference on Mineral Exploration, (ed.) A.G. Gubins; Prospectors and Developers Association of Canada, Toronto, Ontario, p Hajnal, Z., Reilkoff, B., Pandit, B., White, D., Adam, E., Matthews, R., and Koch, R. 2: Seismic modeling prior to the EXTECH-IV Athabasca Basin Seismic Reflection Survey; in Summary of Investigations 2, Volume 2; Saskatchewan Geological Survey, Saskatchewan Energy and Mines, Miscellaneous Report 2-4.2, p Hajnal, Z., Stauffer, M.R., and King, M.S. 1983: Petrophysics of the Athabasca Basin near the Midwest uranium deposit; in Uranium Exploration in Athabasca Basin, Saskatchewan, Canada, (ed.) E.M. Cameron; Geological Survey of Canada, Paper 82-11, p Hajnal, Z., Takács, E., White, D.J., Györfi, I., Powell, B., and Koch, R. 27: Regional seismic signature of the basement and crust beneath the McArthur River mine district, Athabasca Basin, Saskatchewan; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Han, D.-H., Nur, A., and Morgan, D. 1986: Effects of porosity and clay content on wave velocities in sandstones; Geophysics, v. 51, p Hobson, G.D. and MacAuley, H.A. 1969: A seismic reconnaissance survey of the Athabasca Formation, Alberta and Saskatchewan; Geological Survey of Canada, Paper 69-18, p Hoeve, J. and Quirt, D. 1984: Uranium mineralization and host rock alteration in relation to clay mineral diagenesis and evolution of the Mid-Proterozoic Athabasca Basin, Saskatchewan, Canada; Saskatchewan Research Council, Publication R B, 19 p. Hoffman, P.F. 199: Subdivision of the Churchill province and extent of the Trans-Hudson Orogen; in Early Proterozoic Trans-Hudson Orogen of North America, (ed.) J.F. Lewry and M.R. Stauffer; Geological Association of Canada, Special Publication 37, p

25 D.J. White et al. Jamieson, B.W. and Spross, J. 2: The exploration and development of the high grade McArthur River uranium orebody; Fachaufsatze, v. 7/8, 13 p. Jefferson, C.W. and Delaney, G.D. 21: EXTECH IV Athabasca Uranium Multidisciplinary Study: Mid-year 21-2 Overview; in Summary of Investigations 21, Volume 2, part b: EXTECH IV Athabasca Uranium Multidisciplinary Study; Saskatchewan Geological Survey, Saskatchewan Energy and Mines, Miscellaneous Report b, p (CD-ROM). Jefferson, C.W., Thomas, D.J., Gandhi, S.S., Ramaekers, P., Delaney, G., Brisbin, D., Cutts, C., Portella, P., and Olson, R.A. 27: Unconformity-associated uranium deposits of the Athabasca Basin, Saskatchewan and Alberta; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Ji, S., Wang, Q., and Xia, B. 22: Handbook of Seismic Properties of Minerals, Rocks and Ores; Polytechnique International Press, Montreal, Quebec, 63 p. Lebedev, T.S., Korchin, V.A., and Burtnyi, P.A. 1982: Elastic and compliance constants of some rock-forming minerals at high pressure; Geophysical Journal, v. 4, p Lewry, J.F. and Sibbald, T.I.I. 198: Thermotectonic evolution of the Churchill Province in northern Saskatchewan, Canada; Canadian Journal of Earth Sciences, v. 68, p Lydon, J.W. 1995: Sedimentary exhalative sulphides (Sedex); in Geology of Canadian Mineral Deposit Types, (ed.) O.R. Eckstrand, W.D. Sinclair, and R.I. Thorpe; Geological Survey of Canada, Geology of Canada, no. 8, p (also Geological Society of America, The Geology of North America, v. P-1). McGill, B., Marlett, J., Matthews, R., Sopuck, V., Homeniuk, L., and Hubregtse, J. 1993: The P2 North uranium deposit Saskatchewan, Canada; Exploration Mining Geology, v. 2, no. 4, p Milkereit, B., Berrer, E.K., Watts, A., and Roberts, B. 1997: Development of 3-D seismic exploration technology for Ni-Cu deposits, Sudbury Basin; in Geophysics and Geochemistry at the Millennium, Proceedings of Exploration 97: Fourth Decennial International Conference on Mineral Exploration, (ed.) A.G. Gubins; Prospectors and Developers Association of Canada, Toronto, Ontario, p Mwenifumbo, C.J., Elliott, B.E., Drever, G., and Wood, G. 22: Borehole geophysical logging of two deep surface boreholes and one underground borehole at McArthur River; in Summary of Investigations 22, Volume 2; Saskatchewan Geological Survey, Saskatchewan Industry and Resources, Miscellaneous Report , p Mwenifumbo, C.J., Elliott, B.E., Jefferson, C.W., Bernius, G.R., and Pflug, K.A. 24: Physical rock properties from the Athabasca Group: designing geophysical exploration models for unconformity uranium deposits; Journal of Applied Geophysics, v. 55, p Mwenifumbo, C.J., Percival, J.B., Bernius, G.R., Elliott, B., Jefferson, C.W., Wasyliuk, K., and Drever, G. 27: Comparison of geophysical, mineralogical, and stratigraphic attributes in drillholes MAC-218 and RL-88, McArthur River uranium camp, Athabasca Basin, Saskatchewan; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Mwenifumbo, C.J., Pflug, K.A., Elliot, B.E., Jefferson, C.W., Koch, R., Robbins, R., and Matthews, R. 2: Multiparameter borehole geophysical logging at the Shea Creek and McArthur River Projects; parameters for exploration, stratigraphy and high resolution seismic studies; in Summary of Investigations 2, Volume 1; Saskatchewan Geological Survey, Saskatchewan Energy and Mines, Miscellaneous Report 2-4.2, p Mwenifumbo, C.J., Pflug, K.A., Elliott, B.E., Jefferson, C.W., Koch, R., Robbins, J., and Powell, B. 21: Multiparameter borehole geophysical logging at Cluff Lake and McArthur River Projects: new parameters for exploration, stratigraphy and high-resolution seismic studies; in Summary of Investigations 21, Volume 2, Part b: EXTECH IV Athabasca Uranium Multidisciplinary Study; Saskatchewan Geological Survey, Saskatchewan Energy and Mines, Miscellaneous Report b, p (CD-ROM). Ouassaa, K. and Forsyth, D. 22: Interpretation of seismic and potential field data from western New York State and Lake Ontario; Tectonophysics, v. 353, p Overton, A. 1977: Seismic determination of basement depths, Athabasca Basin; in Report of Activities, Part C; Geological Survey of Canada, Paper 77-1C, p Portella, P. and Annesley, I.R. 2a: Paleoproterozoic tectonic evolution of the eastern sub-athabasca basement, northern Saskatchewan: integrated magnetic, gravity, and geological data; extended abstract in GeoCanada 2: the Millennium Geoscience Summit; Joint Meeting of the Canadian Geophysical Union, Canadian Society of Exploration Geophysicists, Canadian Society of Petroleum Geologists, Canadian Well Logging Society, Geological Association of Canada, and the Mineralogical Association of Canada, May 29 June 2, 2, Calgary, Alberta, Canada (also GAC-MAC Program with Abstracts, v. 25), Iron Leaf Communications, Calgary, Alberta (CD-ROM), 647.PDF, 4 p. 2b: Paleoproterozoic thermotectonic evolution of the eastern sub-athabasca basement, northern Saskatchewan: Integrated geophysical and geological data; in Summary of Investigations 2, Volume 2; Saskatchewan Geological Survey, Saskatchewan Energy and Mines, Miscellaneous Report 2-4.2, p Ramaekers, P. 199: Geology of the Athabasca Group (Helikian) in northern Saskatchewan; Saskatchewan Energy and Mines, Saskatchewan Geological Survey, Report 195, 49 p. Ramaekers, P., Jefferson, C.W., Yeo, G.M., Collier, B., Long, D.G.F., Drever, G., McHardy, S., Jiricka, D., Cutts, C., Wheatley, K., Catuneanu, O., Bernier, S., Kupsch, B., and Post, R. 27: Revised geological map and stratigraphy of the Athabasca Group, Saskatchewan and Alberta; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Robert, F. 1995: Quartz-carbonate vein gold; in Geology of Canadian Mineral Deposit Types, (ed) O.R. Eckstrand, W.D. Sinclair, and R.I. Thorpe; Geological Survey of Canada, Geology of Canada, no. 8, p (also Geological society of America, The Geology of North America, v. P-1). Schoenberger, M. and Levin, F.K. 1974: Apparent attenuation due to intrabed multiples; Geophysics, v. 39, p Scott, F. 1983: Midwest Lake uranium discovery; in Uranium Exploration in Athabasca Basin, Saskatchewan, Canada; (ed.) E.M. Cameron; Geological Survey of Canada, Paper 82-11, p Sheriff, R.E. and Geldart, L.P. 1995: Exploration Seismology; Cambridge University Press, Cambridge, United Kingdom, 592 p. Sibbald, T.I.I. 1986: Overview of the Precambrian geology and aspects of the metallogenesis of northern Saskatchewan; in Economic Minerals of Saskatchewan, (ed.) C.F. Gilboy and L.W. Vigrass; Saskatchewan Geological Society, Special Publication 8, p Stewart, R.R., Huddleston, P.D., and Kong Kan, T. 1984: Seismic versus sonic velocities: a vertical seismic profiling study; Geophysics, v. 49, p Stone, D.G. 1994: Designing seismic surveys in two and three dimensions; Geophysical References Series, v. 5, Society of Exploration Geophysicists, Tulsa, Oklahoma, 244 p. 387

26 GSC Bulletin 588 Thomas, D.J., Jefferson, C.W., Yeo, G.M, Card, C., and Sopuck, V.J. 22: I. Introduction; in Trip A1: The Eastern Athabasca Basin and its Uranium Deposits, (ed.) N. Andrade, G. Breton, C.W. Jefferson, D.J. Thomas, G. Tourigny, S. Wilson, and G.M. Yeo; Geological Association of Canada Mineralogical Association of Canada, Field Trip Guidebook, p Thomas, D.J., Matthews, R.B., and Sopuck, V. 2: Athabasca Basin (Canada) unconformity type uranium deposits: exploration model, current mine developments and exploration directions; in Geology and Ore Deposits 2: the Great Basin and Beyond; (ed.) J.K. Cluer, J.G. Price, E.M. Struhsacker, R.F. Hardyman, and C.L. Morris; Geological Society of Nevada Symposium Proceedings, Reno, Nevada, May 15 18, 2, v. 1, p Thomas, M.D. and Wood, G. 27: Geological significance of gravity anomalies in the area of McArthur River uranium deposit, Athabasca Basin, Saskatchewan; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Toohill, K., Siegesmund, S., and Bass, J.D. 1999: Sound velocities and elasticity of cordierite and implications for deep crustal seismic anisotropy; Physics and Chemistry of Minerals, v. 26, p Tosaya, C. and Nur, A. 1982: Effects of diagenesis and clays on compressional velocities in rocks; Geophysical Research Letters, v. 9, p Tourigny, G., Quirt, D.H., Wilson, N.S.F., Wilson, S., Breton, G., and Portella, P. 27: Geological and structural features of the Sue C uranium deposit, McClean Lake area, Saskatchewan; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). Tourigny, G., Wilson, G., Breton, G., and Portella, P. 22: Geology of the Sue C uranium deposit, McClean Lake area, northern Saskatchewan; in Trip A1: The Eastern Athabasca Basin and its Uranium Deposits, (ed.) N. Andrade, G. Breton, C.W. Jefferson, D.J. Thomas, G. Tourigny, S. Wilson, and G.M. Yeo; Geological Association of Canada Mineralogical Association of Canada, Field Trip Guidebook, p Tran, H.T. 21: Tectonic evolution of the Paleoproterozoic Wollaston Group in the Cree Lake zone, northern Saskatchewan, Canada; Ph.D. thesis, University of Regina, Regina, Saskatchewan, 458 p. White, D.J., Hajnal, Z., Reilkoff, B., Jamieson, D., Roberts, B., Koch, R., and Powell, B. 21: EXTECH-IV Subproject 1: preliminary results from the McArthur River 2D high-resolution seismic reflection survey; in Summary of Investigations 21, Volume 2; Saskatchewan Geological Survey, Saskatchewan Energy and Mines, Miscellaneous Report 2-4.2, p Widess, M.B. 1973: How thin is a thin bed?; Geophysics, v. 38, p Wyllie, M.R.J., Gregory, A.R., and Gardner, L.W. 1956: Elastic wave velocities in heterogeneous and porous media; Geophysics, v. 21, p Yeo, G.M., Jefferson, C.W., and Ramaekers, P. 27: Comparison of lower Athabasca Group stratigraphy among depositional systems, Saskatchewan and Alberta; in EXTECH IV: Geology and Uranium EXploration TECHnology of the Proterozoic Athabasca Basin, Saskatchewan and Alberta, (ed.) C.W. Jefferson and G. Delaney; Geological Survey of Canada, Bulletin 588 (also Saskatchewan Geological Society, Special Publication 18; Geological Association of Canada, Mineral Deposits Division, Special Publication 4). 388

27 GSC Bulletin 588 Article by: D.J. White et al. Line B-B' 4 6 Vertical gradient (nt/m) N Northing (km) Line A-A' P2-N 64 P Easting (km) 5 1 km Figure 2. Location map showing the regional seismic lines (A and B) and locations of mineralized zones (P2, P2-N) superposed on a vertical derivative aeromagnetic map of the area. The numbers along the lines correspond to common depth point (CDP) stations (station spacing is 12.5 m). The white rectangle outlines the area shown in Figure 3. The area of the black rectangle corresponds to the black rectangle in Figure 3. 61

28 Article by: D.J. White et al. GSC Bulletin 588 MAC-218 Clay proportions (%) RES ( Ω m) Velocity -1 (km s ) Density -3 (g cm ) 5 Clay intraclasts (mm) Congl (%) MTG (mm) Gamma-ray (cps) Strat. units Depth (m) Full waveform log Time ( µ s) amplitude Tube-wave Casing Kaolin. MFd3 1 Dravite MFd2 15 MFd1 Chlor. 2 MFc2 Illite 25 MFc1 Fracture zones MFb2 MFb1 Dickite RD4 RD3 Plugged Plugged RD2 RD1 RD Bsmt wedge Figure 7. Comparison of stratigraphic units with gamma, grain-size, geophysical, and clay mineral logs for MAC-218, (after Fig. 2 and 4 of Mwenifumbo et al. (27)). Also shown on the right are a full waveform sonic log and tube-wave amplitudes that have been used to identify shallow fracture zones as indicated by the solid black bars. The gamma-ray log has been divided into stratigraphic (Strat.) units: RD = Read Formation (former MFa); MF = Manitou Falls Formation; MFb = Bird Member; MFc = Collins Member; MFd = Dunlop Member (Ramaekers et al., 27). Subscripts on RD, MFb, MFc, and MFd are seismically differentiated subunits within these lithostratigraphic units defined by Yeo et al. (27). Horizontal dashed line is geologically logged silicification front. See text for discussion. MTG = maximum grain size; RES = resistivity; Congl = per cent of 1 m interval that contains conglomerate beds 2 cm or more thick; clay intraclasts = aggregate thickness per metre of tabular clay clasts; clay proportions were determined by in situ spectrometry of matrix clay that typically constitute.5 2.5% of the rock. Kaolin. = kaolinite, Chlor. = chlorite. 62

29 GSC Bulletin 588 Article by: D.J. White et al. 8.2 Reflection a) Vertical: fault and orebody d) Horizontal: fault and orebody 8 8 Position (m) 12 Diffraction Scattering Position (m) 12 Position (m) Surface Multiple 16 Position (m) Boundary Diffraction b) Vertical: fault only e) Horizontal: fault only Position (m) 12 c) Vertical: orebody only f) Horizontal: orebody only P-P P-P P-S P-P P-S P-S Position (m) P-P P-S P-P P-S P-P P-S (db) Figure 11. Simulated 2-D vertical-incidence seismic-reflection image (unmigrated) in the vicinity of the orebody (see Fig. 1 for the orebody model). The approximate location of the orebody (yellow zone) and faulted sandstone-basement interface (white curve) are shown. a), b), c) Vertical component (i.e. what would be measured on vertical component geophones), and d), e), f) horizontal component response; P-P = P-wave diffractions and P-S = S-wave (converted) diffractions. Responses are shown for the complete model (Fig. 11a, d), faulted sandstone-basement interface only (Fig. 11b, e), and for the orebody only (Fig. 11c, f). 63

30 Article by: D.J. White et al. GSC Bulletin 588 A Line Line Approximate depth (m) MAC-259 MAC-138 P2MZ RL-68 MAC-218 Approximate depth (m) S Times from tabulated depths (v = 45 5) Base of 'unconformity reflection' Interpreted fault and/or shear zone RL-46 RL-89 a) b) NW S2 e c b a P2 F1 B Line 12 d?s1 P2 Figure 15. Unmigrated 2-D high-resolution seismic profiles (numbers along top are CDP stations): a) line 12, and b) line 14. The metamorphic basement is interpreted to lie immediately beneath the prominent reflection S2, recognizing that the reflection could be associated with anyone of the scenarios discussed in Variability at the unconformity in the Seismic reflectivity of the basin rocks section. Locations where S2 changes depth abruptly or where there is a clear truncation have been indicated as faults. In some cases, the interpreted fault has been continued into the overlying sandstone column (based on observed discontinuities there) or into the basement where dipping reflections are observed. The two-way traveltime to the unconformity has been plotted for various boreholes by converting depth to time using the sonic or VSP interval velocities. The two horizontal lines on each well bore indicate the variation in the two depth estimates. The grey rectangles indicate regions of the data that are shown in Figure 19 at an expanded scale. Labels are referred to in the text. SE S1 S2 64

31 GSC Bulletin 588 Article by: D.J. White et al. Line 12: S/N (db) S/N (db) 9 2 Receiver station NW Sequential shot number Figure 16. Signal-to-noise ratios (S/N) along line 12 plotted for each shot-point versus receiver station. Noise and signal levels were determined in a 3 ms window prior to and following the first arrival, respectively. 65

32 Article by: D.J. White et al. GSC Bulletin 588 a) Line 12: SG 745 Channel number (db) GR FB SFB S b) Line 14: SG 773 Channel number S SFB.5 FB GR Figure 18. True absolute amplitude representation of example shot gathers from lines 12 and 14. Amplitudes are determined for the raw gathers. Refraction statics have been applied. Note the different scales for lines 12 and 14, although the range is 4 db in both cases. FB = first breaks, SFB = shear-wave first arrival, S2 = unconformity reflection, GR = ground roll. 66

33 GSC Bulletin 588 Article by: D.J. White et al. Figure 22. Three-dimensional survey geometry. The black rectangle surrounding the McArthur River mining camp is the same area as the black rectangle in Figure 3. Three-dimensional stack fold diagram showing the number of seismic traces that fall within individual 15 x 15 m bins. A seismic trace represents the data recorded at a single geophone station from a single shot point. Each seismic trace is assigned to a bin in the grid based on the location of the midpoint between the source and geophone for that trace. Multiple traces (i.e. multifold) falling within a given bin are added together to increase the signal-to-noise ratio of the data that ultimately form a 3-D image. Also shown are the approximate locations of the McArthur River ore zones (yellow ellipses), and the locations of the vertical slices through the 3-D data volume (3-D-1, 3-D-2 of Fig. 23). 67

34 Article by: D.J. White et al. GSC Bulletin 588 Common depth point number a) NW NW Vp (m/s) b) S2 S2 F2 F3 S2 High velocity zone Figure 25 P2 F1 B Common depth point number F2 F3 S2 P2 F1 B Figure 24. The upper 2 s of the line B-B' profile: a) unmigrated, and b) time-migrated section. The colour overlay depicts subsurface velocities determined from inversion of the first-arrival traveltimes for line B-B'; F1, F2, F3, F4 indicate the locations of interpreted basement faults. Other labels are defined in the text. F4 F4 SE S2 SE 68