Subsurface mapping in the Iberian Pyrite Belt using seismic reflection profiling and potential-field data

Abstract

The Iberian Pyrite Belt (IPB) hosts world-class massive sulphide deposits, such as Neves-Corvo in Portugal and Rio Tinto in Spain. In Portugal, the Palaeozoic Volcanic-Sedimentary Complex (VSC) hosts these ore deposits, extending from the Grândola-Alcácer region to the Spanish border with a NW-SE to WNW-ESE trend. In the study area, between the Neves-Corvo mine region and Alcoutim (close to the Spanish border), the VSC outcrops only in a small horst near Alcoutim. Sparse exploration drill-hole data indicate that the depth to the top of the VSC varies from several 100 m to about 1 km beneath the Mértola Formation Flysch cover. Mapping of the VSC to the SE of Neves-Corvo mine is an important exploration goal and motivated the acquisition of six 2D seismic reflection profiles with a total length of approximately 82 km in order to map the hidden extension of the VSC. The data, providing information deeper than 10 km at some locations, were integrated in a 3D software environment along with potential-field, geological and drill-hole data to form a 3D structural framework model. Seismic data show strong reflections that represent several long Variscan thrust planes that smoothly dip to the NNE. Outcropping and previously unknown Late Variscan near-vertical faults were also mapped. Our data strongly suggest that the structural framework of Neves-Corvo extends south-eastwards to Alcoutim. Furthermore, the VSC top is located at depths that show the existence within the IPB of new areas with good potential to develop exploration projects envisaging the discovery of massive sulphide deposits of the Neves-Corvo type.

http://crossmark.crossref.org/dialog/?doi=10.1007/s00531-016-1340-1&domain=pdf

Web End = Int J Earth Sci (Geol Rundsch) (2017) 106:899915 DOI 10.1007/s00531-016-1340-1

ORIGINAL PAPER

http://crossmark.crossref.org/dialog/?doi=10.1007/s00531-016-1340-1&domain=pdf

Web End = Subsurface mapping in the Iberian Pyrite Belt using seismic reection proling and potentialeld data

Joo Carvalho1 Carlos Inverno1 Joo Xavier Matos1 Carlos Rosa1,4 Isabel Granado1 Tim Branch2 Patrcia Represas1 Livia Carabaneanu2 Lus Matias3 Pedro Sousa1

Abstract The Iberian Pyrite Belt (IPB) hosts world-class massive sulphide deposits, such as Neves-Corvo in Portugal and Rio Tinto in Spain. In Portugal, the Palaeozoic Volcanic-Sedimentary Complex (VSC) hosts these ore deposits, extending from the Grndola-Alccer region to the Spanish border with a NWSE to WNWESE trend. In the study area, between the Neves-Corvo mine region and Alcoutim (close to the Spanish border), the VSC out-crops only in a small horst near Alcoutim. Sparse exploration drill-hole data indicate that the depth to the top of the VSC varies from several 100 m to about 1 km beneath the Mrtola Formation Flysch cover. Mapping of the VSC to the SE of Neves-Corvo mine is an important exploration goal and motivated the acquisition of six 2D seismic reection proles with a total length of approximately 82 km in order to map the hidden extension of the VSC. The data, providing information deeper than 10 km at some

Received: 9 January 2015 / Accepted: 22 May 2016 / Published online: 15 June 2016 Springer-Verlag Berlin Heidelberg 2016

locations, were integrated in a 3D software environment along with potential-eld, geological and drill-hole data to form a 3D structural framework model. Seismic data show strong reections that represent several long Variscan thrust planes that smoothly dip to the NNE. Outcropping and previously unknown Late Variscan near-vertical faults were also mapped. Our data strongly suggest that the structural framework of Neves-Corvo extends south-eastwards to Alcoutim. Furthermore, the VSC top is located at depths that show the existence within the IPB of new areas with good potential to develop exploration projects envisaging the discovery of massive sulphide deposits of the Neves-Corvo type.

Keywords Iberian Pyrite Belt Massive sulphide deposits Volcanic-Sedimentary Complex Seismic reection Magnetic anomalies Bouguer anomalies

* Joo Carvalho [email protected]

Carlos Inverno [email protected]

Joo Xavier Matos [email protected]

Carlos Rosa [email protected]

Isabel Granado [email protected]

Tim Branch [email protected]

Patrcia Represas [email protected]

Livia Carabaneanu [email protected]

Lus Matias [email protected]

Pedro Sousa [email protected]

1 Laboratrio Nacional de Energia e Geologia, Apartado

7586-Alfragide, 2710-966 Amadora, Portugal

2 Prospectiuni S.A., Strada Caransebe 1, Bucureti,

012271 Bucarest, Romania

3 Instituto Dom Luiz, Faculdade de Cincias, Universidade de

Lisboa, Campo Grande, 1749-016 Lisbon, Portugal

4 Present Address: Empresa de Desenvolvimento Mineiro (EDM),

Rua Sampaio e Pina, 1 3.Dto., 1070-248 Lisbon, Portugal

1 3

900 Int J Earth Sci (Geol Rundsch) (2017) 106:899915

Introduction

The knowledge of the Variscan structure with depth is essential for ore exploration in the Iberian Pyrite Belt (IPB) area. Metallic ore deposits, such as the massive sulphide deposits of the IPB, are often associated with dipping (30 60) structures. Strong deformation and metamorphism limit the use of seismic methods to locate them. Reection seismic methods had long been employed to study the tectonics and stratigraphy of sedimentary basins, where acoustic impedance contrasts are generally strong enough to allow mapping of the subsurface geological features. Recent developments in migration algorithms, signal-to-noise ratio (SNR) enhancement, such as in the common-reection surface (CRS) method can improve the mapping of such deposits in structurally complex environments (e.g. Mann et al. 1999, 2007; Hubral 1999). 3D seismic proling and advances in electronics such as 24-bit A/D converters have also contributed to improving the utility of reection seismic methods in hard-rock environments (e.g. Cordsen et al. 2000; Vermeer 2002; Eaton 2003).

In the last decades, the seismic reection method has experienced a dramatic increase in the number of applications to hard-rock exploration (e.g. Kim et al. 1994; Radzevicius and Pavlis 1999; Juhlin et al. 2002; Juhlin and Stephens 2006; Eaton 2003; Schmelzbach et al. 2007; Malehmir et al. 2012a; Heinonen et al. 2013; Ehsan et al. 2014). The mining sector is one of the elds where the seismic reection method has shown great utility to complement the more generally applied gravimetric, magnetic, electrical and electromagnetic methods. When combined with traditional methods, reection seismic method can provide structural information in complex geological environments (Manzi et al. 2012, 2013).

Both 2D and 3D seismic methods have now provided successful case histories in the mining industry (Adam et al. 1998; Calvert and Li 1999; Ayarza et al. 2000; Cheraghi et al. 2011, 2012; Malehmir et al. 2011, 2012a, b; Urocsevic et al. 2012; Ahmadi et al. 2013). On the Portugal mainland, the rst seismic reection proles acquired in Alentejo for this purpose were supported by Somincor, the mining company operating in the Neves-Corvo mine, in 1991 and 1996 (Carvalho et al. 1996). Twelve proles were shot by Compagnie Gnrale de Gophysique (CGG) using explosives as a seismic source and a system of 48 or 24 active channels. At the time, the results were of poor quality and were not very useful to the local mining companies and no further surveys were acquired during the following years.

Several years after the last CGG survey in Portugal, the rst deep seismic reection proling started in the Spanish part of the IPB, close to the Portuguese border. For the rst time, the whole crust down to Moho was imaged in the

area, giving unprecedented information about the rooting of several outcropping structures and geological formations (e.g. Simancas et al. 2003; Carbonell et al. 2004). In Portugal, useful results were obtained using seismic reection and gravimetric modelling for shallow massive sulphides exploration in the western section of the IPB. Geophysical data were used to dene the depth to the Palaeozoic basement beneath the Cenozoic sedimentary cover (Oliveira et al. 1998; Carvalho et al. 2011).

Encouraged by these results, under the scope of the EU Seventh Framework Programme (FP7)-funded Promine project, it was decided to acquire a set of 2D seismic reection proles to provide structural information down to 45 km in the SE sector of the Portuguese IPB region (Inverno et al. 2015a). The aim was to determine whether the structural framework associated with the Neves-Corvo mine, together with a lineament of Late Variscan vein-type Cu occurrences, extends SE to the Spanish border, south of Alcoutim (Carvalho 1991; Matos et al. 2003; Reiser et al. 2011), as suggested by several holes drilled in the ysch cover represented by the Mrtola Formation shales and greywackes of Upper Visean age (Oliveira et al. 2004). These data would also help to outline the upper and lower limits of the Volcanic-Sedimentary Complex (VSC) in the region. The use of seismic proles could also allow acquisition of information about the IPB basement, which at the Neves-Corvo region is constituted by the PhylliteQuartzite Group (PQG) of Famennian age (Oliveira et al. 2013). Therefore, in the scope of this work, seismic reection data were acquired, processed and interpreted together with geological 1/200 000 scale mapping, drill-hole data and information from different surveys: aeroradiometric (gamma-ray total counts), land gravimetric and aeromagnetic data in a 3D modelling environment.

The seismic reection method is the most suited for the purpose of geological mapping of the IPB due to its superior resolution and depth penetration. Drill-holes usually do not exceed a depth of 2 km. The available gravimetric data have provided very valuable results for the discovery of several important mineral deposits such as Neves-Corvo. However, not only are these data affected by the traditional non-uniqueness problems associated with inversion (e.g. Menke 2012), but also generally do not provide useful information below a depth of 12 km. Electrical and electromagnetic methods, often applied in the area, are useful for locating ore bodies down to 12 km, but they do not provide detailed structural information, such as the location of major thrust planes and faults.

The knowledge of the Variscan structure with depth is essential for ore exploration in the IPB area. Models of the SPZ were until recently based on geological mapping and drill-hole data (not over 2 km depth). The acquisition of the IBERSEIS (e.g. Simancas et al. 2003) and ALCUDIA (e.g.

1 3

Int J Earth Sci (Geol Rundsch) (2017) 106:899915

901

Martnez-Poyatos et al. 2012) deep seismic reection proles in Spain close to the Portuguese border as a part of the SW Iberia EUROPROBE project have provided for the rst time valuable information on the Iberian Variscan basement at depth.

Geological setting

The South Portuguese Zone (SPZ) is the southernmost part of the Iberian massif on the south-western Iberian Peninsula and is one of the best exposed fragments of the Vari-scan orogeny in Western Europe (e.g. Matte 1986). During Late DevonianEarly Carboniferous time, magmatic activity was intense within the SPZ with occurrence of submarine bimodal volcanism that resulted in the generation of world-class massive sulphide deposits (Carvalho et al. 1999; Sez et al. 1999). This volcanogenic province is usually designated as the Iberian Pyrite Belt (IPB).

The study area within the IPB is located south of Castro Verde, in southern Alentejo Province, south Portugal. It is adjacent to Neves-Corvo mine and extends to the SE to

the Spanish border (Fig. 1). The study area is part of the (Middle) Late DevonianEarly Carboniferous IPB succession. This succession is stratigraphically composed of two main distinct sequences (Boogaard 1967; Schermer-horn 1971; Carvalho et al. 1976): the Upper Famennian Upper Visean Volcanic-Sedimentary Complex (VSC) and the Lower GivetianUpper Famennian PhylliteQuartzite Group (PQG; ages in Pereira et al. 2008; Oliveira et al. 2013; Matos et al. 2014; Fig. 1). The VSC is composed of felsic volcanics, basic volcanics, black shales, siltstones, volcanogenic sedimentary rocks, cherts and jaspers.

The IPB sequence is stratigraphically overlain by marine sedimentary rocks of the Baixo Alentejo Flysch Group (BAFG), represented in the study area by the Mrtola Formation (Upper Visean) and by the Mira Formation (Upper ViseanSerpukhovian; Carvalho et al. 1976; Oliveira 1983; Matos et al. 2003; Oliveira et al. 2004; Pereira et al. 2008). A considerable amount of palynological, ammonoid and geochronological ages have been determined on these units. The Mrtola Formation at depth was recognized by exploration drill-holes, showing variable thickness, locally >1 km. At Neves-Corvo (see location in Fig. 1) and NW of

Fig. 1 Geological sketch (after Oliveira et al. 1992) and location of the study area, which is indicated by the dashed line

1 3

902 Int J Earth Sci (Geol Rundsch) (2017) 106:899915

the Alcoutim VSC horst structure, the Mrtola Formation is present in duplex tectonic structures dominated by thrusts, located between VSC units. North of the IPB, different metasedimentary units of Lower DevonianUpper Famennian age crop out, namely the Pulo do Lobo Formation, in between the Chana Group and the Ferreira-Ficalho Group (Fig. 1; Oliveira et al. 2004).

The study area is located between the Neves-Corvo mine (Fig. 1), in the SE sector of the Neves-CorvoRosrio antiform, and the Spanish border, located 50 km to the south-east (Fig. 1). It comprises four main antiformal structures, Castro Verde-Casvel, Neves Corvo-Rosrio, Panias-Ourique, and possibly Alcoutim (see location in Fig. 1). These structures are all placed along distinct volcanic axes. The exposed VSC lithologies are dominant in these antiforms. The PQG is only exposed locally at the SE area of the Rosrio antiform (Oliveira et al. 2013) and at depth in the Neves-Corvo mine (Relvas et al. 2006; Pereira et al. 2008).

The geological sequence is represented from base to top by the PQG (>100 m thick), of Upper Famennian to Strunian age, that forms the IPB siliciclastic lower group and consists of shales, siltstones and quartzites, with limestones on top of the succession; and the VSC, which starts with a Famennian lower VSC sequence (500 m thick), with rhyolites, rhyodacites, volcaniclastic sedimentary rocks, shales, basalts and dolerite sills (Oliveira et al. 2013). The Neves-Corvo sulphide ores are associated in this lower VSC sequence with black shales of the Neves Formation, of Strunian age (Pereira et al. 2008), and felsic volcanics and are represented by seven massive sulphide lenses and related stockworks.

In the mine area, the lower VSC is unconformably over-lain by the Visean Mrtola Fm. (Mt2) greywackes and shales (up to several decametres thick; Pereira et al. 2008; Oliveira et al. 2013). The Upper Visean VSC sequence (450 m thick) is present, after a Tournaisian gap, on top

of those turbidites. Shales, such as grey to black shales, purple shales, and shales with phosphate nodules, and chert, ne volcanogenic sedimentary rocks and felsic volcanic rocks, are part of this sequence. The upper VSC sequence is overlain by the Visean Mrtola Fm. (Mt1; older than Mt2) [>1000 m thick], dominant in the study area, with greywackes, siltstones and shales. In the SW part of the study area, the younger Mira Fm. (10002500 m thick) occurs SW of the Mrtola Fm., with ne-grained greywackes, shales and siltstones. Within the study area, the VSC outcrops only at the SE area of the Rosrio Neves-Corvo structure and in the Alcoutim horst (Fig. 1; Leca 1976; Leca et al. 1983; Oliveira et al. 2013). A thrust between the upper VSC and Mt2 (ysch) is a distinctive guide in the mine area to the Neves-Corvo ore horizon which is represented by the black shales of the Neves Formation and felsic volcanics (Carvalho 1991).

All these stratigraphic units are affected by Variscan NW-trending folds and associated (main) cleavage dipping 6070 to the NE, with vergence towards SW (Oliveira et al. 2004). The Neves-CorvoRosrio antiform (VSC PQG) plunges at low angle to the SE under the exposed Mrtola Fm. rocks, as indicated by a few drill-holes and different types of geophysical surveys. Several thrust faults parallel to the main cleavage occur, a few extending continuously throughout the study area. The upper, and in places, the lower VSC sequences are commonly stacked in tectonic sheets that were thrusted over the structurally lower Mrtola Formation rocks. There are also Late Variscan near-vertical faults in two main families: NWSE, dextral, and NESW, sinistral, both in places with small deposits and occurrences mostly of Cu, but also of Ba(Pb) and Sb(AU) (Matos et al. 2003; Reiser et al. 2011; Oliveira et al. 2004, 2013; Inverno et al. 2015b, see location in Matos and Filipe 2013). MnFe small deposits related to VSC upper sequence lithologies are known at NW Neves-Corvo and Alcoutim (Leca 1976; Leca et al. 1983; Matos and Filipe 2013).

Geophysical methods

2D seismic reection data

Acquisition

As stated above, several seismic proles were acquired by CGG in 1991 and 1996 (Carvalho et al. 1996). When interpreted in a GIS environment integrated with drill-hole, geological mapping and other geophysical data, their input suggested that important structural information could be retrieved from the seismic data. Therefore, six 2D seismic proles, approximately NESW oriented, with a total length of 82 km were planned for the area between the Neves-Corvo mine and the Portuguese/Spanish border (Fig. 1). The location of the seismic proles was chosen to investigate the crust below the ysch cover for the possible extension of the Neves-Corvo volcanic axis to the Alcoutim area (close to the Spanish border), where the VSC outcrops again, and its southern region; along this extension, the VSC rocks were intersected by exploration drill-holes at 400810 m depth (Billinton exploration project developed in the 1980s). Figure 1 shows a geological map with the location of the six seismic proles.

Due to the required depth of penetration (45 km), either explosives or heavy vibrators are needed to be used as seismic sources. Permits, licenses and the invasive nature of explosives led altogether to the choice of using vibrators. Three synchronized ION AHV-IV vibroseis trucks of 28.5 tonnes each were therefore used as seismic source. A linear

1 3

Int J Earth Sci (Geol Rundsch) (2017) 106:899915

903

Table 1 Acquisition parameters of the Promine seismic reection proles used in this study

Seismic source Sweep freq. (Hz) Offset range (m) Active channels

Channel spacing (m)

Receivers per channel (linear array)

Source move-up (m)

Nominal CMP fold

Three synchronized 28.5-tonne each vibrators

1090 2512,000 480 25 6 50 120

Table 2 Major steps of the different processing ows applied to the Promine seismic reection proles used in this study

Text in italics indicates PSTM sequence mentioned in ow 3 and text, while text in bold represents CMP stack sequence referred to in ow 4 and text

Flow 1 Flow 2 Flow 3 Flow 4

CMP/crooked line geometry CMP/crooked line geometry CMP/crooked line geometry CMP/crooked line geometry Trace editing + freq. ltering Trace editing + freq. ltering Trace editing + freq. ltering Trace editing + freq. ltering

Muting Muting Muting MutingRefraction statics Refraction statics Refraction statics Refraction staticsFlat datum conversion Flat datum conversion Flat datum conversion Flat datum conversion FK ltering FK ltering FK ltering FK lteringSpiking deconvolution Spiking deconvolution Spiking deconvolution Spiking deconvolution Velocity analysis Velocity analysis Velocity analysis Velocity analysis Residual statics Residual statics Residual statics Residual statics

Final datum correction + velocity analysis NMO + DMO + stack FX decon + CRS geometry FX decon + CRS geometry

Kirchhoff PSTM Dip Explicit FD Time Migration PSTM sequence CMP stack sequence

FX ltering + AGC FX ltering + AGC

sweep of 16 s was used with a frequency varying between 10 and 90 Hz to allow recording shallow and deeper reections. Two or three vibrations per vibration point were carried out.

The data acquisition system used was a Sercel 428 E-Unite Full. SG 10-Hz receivers were used to record the seismic wave eld, which was sampled at a 2-ms rate for 6 s. The acquisition parameters are presented in Table 1. The use of a wireless system allowed a superior exibility in data acquisition. Villages, roads, small rivers could be easily crossed, and the initial design of the proles could be followed more closely.

Taking into consideration the actual frequency content of the recorded data and the velocities obtained during data processing (see next section), we can expect, using the Rayleigh criterion (Born and Wolf 1980), a maximum vertical resolution of approximately 12 m at shallow depths and of around 150 m at large depths. Horizontal resolution, given by the rst Fresnel zone, is approximately 100 m and 1 km for depths of 0.5 km and of 4 km, respectively.

Data processing

Several processing sequences were attempted for each prole: a standard processing sequence, with pre-stack time

migration (PSTM) and post-stack migration, and a second processing sequence, with the common-reection surface (CRS) method (Mann et al. 1999), and followed again by PSTM and post-stack migration.

All processing sequences, listed in Table 2, were similar until the application of residual statics. From this point onwards, the standard and PSTM processing sequences diverged. The CDP gathers were either NMO (normal move-out) and DMO (dip move-out) corrected, stacked and post-stack processed with F-X deconvolution plus AGC applied, and/or prepared for PSTM. In the former case, Steep Dip Explicit Finite Difference Time Migration was applied to the stacked sections, whereas in the latter case, the CDP gathers were prepared for PSTM. Subsequently, a correction to the nal datum was applied, a detailed velocity analysis undertaken, and a 2D Kirchhoff time migration algorithm with an aperture of 35 km was applied to the gathers. To conclude the processing sequence, FK lters and AGC followed and nally, CDP stack. F-X deconvolution and AGC were applied to the stacks.

The CRS processing sequence was identical to the previous one till the pre-PSTM gathers and before NMO correction. After the application of residual statics to the gathers, the following steps were taken: F-X deconvolution, AGC, 2D CRS ZO search and 2D CRS stack. From this

1 3

904 Int J Earth Sci (Geol Rundsch) (2017) 106:899915

point onwards, the sequence was different for standard and PSTM sequences. NMO correction and CDP stack (applying statics to nal datum) followed by Steep Dip Explicit FD Time Migration for the standard sequence and, for the PSTM ow, corrections to nal datum, velocity analysis, PSTM, FK lters and nally stacking. Post-stack processing included F-X deconvolution ltering and AGC (Table 2).

DMO and migration were essential to improve the imaging of steeply dipping features in the data (e.g. Malehmir and Juhlin 2010; Cheraghi et al. 2012). The CRS stacked sections generally resulted in a clear improvement of the signal-to-noise ratio and a better continuity of the seismic reectors. However, for line 3 (Fig. 2), the CRS stack increased the presence of conicting dips, probably due to the structural lateral variations in the region and/or an inappropriate velocity function. CRS stacking is particularly sensitive to dip and curvature of the reectors (Mann et al. 2007), and the local 3D structure might cause additional difculties in performing an adequate velocity analysis. Figure 2 shows the non-migrated CRS and standard CMP stacks for this line, as well as the post-stack migrated and PSTM stacks.

PSTM sections minimized the effects of conicting dips (see Fig. 2) in the data that resulted from the 3D nature of the geological structures, but PSTM often introduced probable artefacts (see Fig. 2) in the stacked sections. These artefacts were identied by visual inspection of non-migrated, PSTM and post-stack migrated sections and also taking into account the available geological data. This is also highlighted in Fig. 2 as line 3, but line 6 presents the clearest example. For the nal interpretation, it was an advantage to have access to processing sequences and the non-migrated stacks.

Potentialeld data

The interpretation of the seismic proles was complemented with available land gravimetric and aeromagnetic data. In spite of their lower resolution when compared to seismic reection data, potential-eld data can provide valuable information. Magnetic susceptibilities are highly sensitive to lithological variations and can indicate depth variations of VSC units that generally correspond to higher total eld magnetic intensities than the surrounding ysch cover, in particular the basic volcanics embedded in the VSC. Density information available from drill-hole data together with gravimetric interpretation can provide information about the presence at depth of denser units such as the PQG, or the possible presence of ore deposits within the VSC. This information is of major importance since seismic data sometimes fail to provide continuous markers between stratigraphic boundaries in hard-rock

environments, due to high deformation, overburden compaction and metamorphism.

An unofcial national Bouguer anomaly map based on the National Gravimetric Network (NGN, http://www.igeo.pt/produtos/Geodesia/gravimetria.htm

Web End =http://www. http://www.igeo.pt/produtos/Geodesia/gravimetria.htm

Web End =igeo.pt/produtos/Geodesia/gravimetria.htm ) is available for Portugal mainland. The NGN is constituted by over 6500 points and two absolute stations, one in northern Portugal and the other in southern Portugal, resulting in a minimum density coverage of 1 data point per 2530 km2. However, several local gravimetric surveys have been carried out in the IPB by the National Laboratory for Energy and Geology (LNEG) and its predecessors in the last decades. These land surveys that present a much higher data point density than the NGN have been acquired since the 1960s using different equipment and needed to be homogenized before a nal Bouguer anomaly map could be completed. Figure 3 shows the Bouguer anomaly map with the location of the seismic lines overlaid. A reference density of 2.65 g/cm3

was used to produce the Bouguer anomaly map.

In 1991, Geoterrex conducted a magnetic survey over the Portuguese sector of the Iberian Pyrite Belt for the Rio Tinto mining company using a caesium vapour magnetometer at a constant ight altitude of 90 m (Carvalho 1995; Carvalho et al. 2012). The ight lines had a SWNE orientation and were generally spaced at 500 m, sometimes 250 m, with perpendicular tie lines. The sampling was carried out every 7.5 m and, taking into consideration noise levels, instrument precision, repeatability, etc., a resolution of approximately 0.1 nT was achieved. The survey, which covers the whole study area, was processed by Geoterrex and is used in this work. We have extracted the IGRF from the magnetic total eld intensity data and reduced it to the pole. The results are shown in Fig. 4.

The coincidence of geological bodies and man-made structures with gravimetric and magnetic anomalies (see examples indicated by letters in Figs. 3, 4) suggests that at least working at this scale, the reduction to the pole placed the magnetic anomalies at their appropriate locations.

Integrated interpretation and geological modelling

Due to that the seismic lines were a few kilometres apart, a seismic interpretation package was not used to pick the seismic horizons. All the seismic and other geophysical information, as well as geological mapping and drill-hole data, was integrated within a 3D software environment. It was possible to validate the data in a progressive way, which was important to detect inconsistencies.

Figures 5, 6, 7 and 8 display the nal interpretation of seismic proles 1, 2, 4 and 6. The Bouguer anomaly and reduced-to-pole magnetic total eld intensity are plotted at the top of the seismic proles. The stacked sections are

1 3

Int J Earth Sci (Geol Rundsch) (2017) 106:899915

905

Fig. 2 Stacked sections resulting from the different processing ows applied to seismic reection prole 3. Top left standard CMP stack; top right CRS stack; bottom far left PSTM CRS stack; bottom left CRS stack with post-stack migration applied; bottom right PSTM CMP stack; bottom far right CMP stack with post-stack migration

applied. Ovals indicate areas of increased signal-to-noise ratio but also of conicting dips due to CRS stacking, while rectangles show zones of probable artefacts introduced by PSTM and where post-stack migration performs better (see text)

shown with the CRS without migration, and with post-stack time migration, producing much better results than PSTM. All the seismic sections extend to the depths of interest of over 4 km, with the south-eastern lines presenting reection events down to 10 km. The stacking velocities from the processing were used for depth conversion. These velocities are in agreement with velocity information from drill-holes located in the region of the Neves-Corvo mine (Yavuz et al. 2015) and other drill-holes that reached the Palaeozoic basement in the Lusitanian Basin (GPEP 1986) and also from refraction proles (Matias 1996).

Three drill-holes, two located 45 m and one about 1 km off line 1 (see Fig. 1) and about 50 m off line 6, intersected a seismic prole. Inspecting for instance prole 1 in detail, drill-hole CT08001, the closest to prole 1, reached a drilled length of 1.89 km and was drilled in 2008 by the

AGC/Lundin Mining Company, while drill-hole CT01 had been previously drilled (1.7 km drilled length) by Somincor in 1992. Figure 9 shows the so-called Cotovio geological cross-section based on drill-holes CT08001 and CT01, geological surface data and drill-hole data in the area. To perform the cross-section, CT08001 drill-hole core was re-logged and studied. A careful data interpretation of the CT01 drill-hole was done, based on the 1990s Somincor technical reports. The Cotovio sector geological data were correlated with the information about the geology of the Neves-Corvo mine SE region. Detailed drill-hole geological log data were used in the cross-section and correlated with the seismic prole interpretation. This cross-section has been overlaid with a degree of transparency over the interpreted stacked section of the recently acquired Promine seismic reection (prole 1) shown in Fig. 5, showing

1 3

906 Int J Earth Sci (Geol Rundsch) (2017) 106:899915

Fig. 3 Bouguer anomalymap available for the study area. The locations of the seismic reection proles and available drill-holes are also displayed, together with surface geology contacts and faults (after Oliveira et al. 1992). Coordinates are shown in the HayfordGauss system, datum 73, in metres. Letters indicate coincidence of gravimetricand magnetic anomalies of the Neves-Corvo mine structure(a) and volcanics of the VSC detected by drilling (b), suggesting that reduction to the pole correctly placed magnetic anomalies

Fig. 4 Total-eld, reduced-to-pole aeromagnetic map of the study area, overlaid bythe location of the seismic reection proles, drill-holes, surface geology contacts and faults (after Oliveira et al. 1992). The area covered by gravimetric data shown in Fig. 3 is indicated. Coordinates are shown in the HayfordGauss system, datum 73, in metres. Letters indicate coincidenceof gravimetric and magnetic anomalies of the Neves-Corvo mine structure (a) and volcanics of the VSC detected by drilling (b), suggesting that reductionto the pole correctly placed the magnetic anomalies

the seismic/drill-hole tie for prole 1. Once geophysical logs are not available for drill-holes CT08001 and CT01, depth conversion of the seismic horizons was carried out using the seismic data processing velocities. The principal seismic reectors of prole 1 are also overlaid in Fig. 9. The consistent correlation between seismic data and drill-hole data in the upper part of the prole is represented by the presence of planar reectors gently dipping to the north-east, corresponding to the boundary between Palaeozoic sedimentary and volcanic rocks.

The Cotovio cross-section, a result of one of the exploration research projects in Iberian Pyrite Belt, provided stratigraphic information down to a depth of 1.6 km in the Cotovio gravity anomaly, located 4 km SE of the Neves-Corvo mine. In both drill-holes (Figs. 5, 9), Baixo Alentejo Flysch and VSC rocks were drilled. In detail, and considering the CT08001 geological log data, the sequence is characterized by shales and greywackes of the Mrtola Fm., intersected until 580 m, followed by VSC sedimentary rocks until 1.06 km. Between 1.06 and 1.6 km, felsic

1 3

Int J Earth Sci (Geol Rundsch) (2017) 106:899915

907

Fig. 5 Gravimetric and magnetic data (a) plotted over CRS stack for seismic prole 1 without (b) and with post-stack time migration with overlaid stratigraphic and structural interpretation (c). The locations of drill-holes CT01 and CT08001 are shown (blue line), and a geological cross-section based on these two drill-holes and previously acquired seismic reection data is overlaid by the interpreted stacked section (see details in Fig. 9). Mmagnetic eld intensity (reduced to the pole), Ba Bouguer anomaly. Black lines faults, yellow lines/T thrust planes, VSC Volcanic-Sedimentary Complex, PQG Phyllite Quartzite Group, red line pre-Devonian horizon, orange lines deep crustal reectors (see text)

Fig. 6 Gravimetric and magnetic data (a) plotted over CRS stack for seismic prole 2 without (b) and with post-stack time migration with overlaid stratigraphic and structural interpretation (c). Mmagnetic eld intensity (reduced-to-pole), Ba Bouguer anomaly. Black lines faults, yellow lines/T thrust planes, green line/VSC Volcanic-Sedimentary Complex, brown line/PQG PhylliteQuartzite Group, red line pre-Devonian horizon, orange lines deep crustal reectors (see text)

volcanics (rhyolite and rhyodacite) were intersected, representing a large volcanic centre, probably a dome structure. Metric intercalation of VSC sedimentary rocks was observed at 1.52 km, corresponding to a tectonic duplex structure.

The CT01 drill-hole log shows two intersections of shales and greywackes of the Mrtola Fm. from 1.02 to 1.08 km and from 1.16 to 1.17 km and reects the presence of important thrust structures, similar to the Neves-Corvo thrust (Oliveira et al. 2004; Inverno et al. 2015b). These fault plans are related to a strong vergence to SW. None of the drill-holes intersects sulphide mineralization (either massive or semi-massive). Neither do the volcanic rocks show important hydrothermal alteration.

Visible reector planes (at about 0.2 s) are related to the Mrtola Flysch/VSC contact and the VSC sedimentary/volcanic rocks contact (Fig. 5). This surface can be correlated with the top of the Cotovio volcanic structure. The reectors related to stratigraphy do not show a good lateral continuity along the stacked section of prole 1.

1 3

908 Int J Earth Sci (Geol Rundsch) (2017) 106:899915

Fig. 7 Gravimetric and magnetic data (a) plotted over CRS stack for seismic prole 4 without (b) and with PSTM applied, overlaid with stratigraphic and structural interpretation (c). Mmagnetic eld intensity (reduced-to-pole), Ba Bouguer anomaly. Black lines faults, yellow lines/T thrust planes, green line/VSC Volcanic-Sedimentary Complex, brown line/PQG PhylliteQuartzite Group, red line pre-Devonian horizon

Fig. 8 Gravimetric and magnetic data (a) plotted over CRS stack for seismic prole 6 without (b) and with post-stack time migration with overlaid stratigraphic and structural interpretation (c). Mmagnetic eld intensity (reduced-to-pole), Ba Bouguer anomaly. Black lines faults, yellow lines/T thrust planes, green line/VSC Volcanic-Sedimentary Complex, brown line/PQG PhylliteQuartzite Group, red line pre-Devonian horizon

Recent work carried out in the region of the Neves-Corvo mine (Yavuz et al. 2015), which included the building up of an elastic properties database with measurements on nearly 400 core samples, a modelling study using full-waveform sonic and pseudo-log data to create synthetic seismic data and subsequent tie to 3D seismic reection data, shows us that the contact between the Mrtola Flysch and the VSC is not a prominent reector compared to intra-ysch reectors or other stratigraphy-related reectors. Furthermore, the work of Yavuz et al. (2015) shows us that the strongest amplitude reectors are related to the presence of massive sulphide deposits (drill-holes PSJ50 and PSK50, in Yavuz et al. 2015). The Mrtola Flysch/VSC contact produces at the best (shale/rhyolite contact in drill-hole SZ24A; Yavuz et al. 2015) a much lower reectivity than massive ore. In our opinion, the ysch/VSC contact reectivity at drill-hole SZ24A is enhanced by the strati-graphic inversion related to a thrust fault. This reinforces

the importance of the structural control in the reectivity observed in the seismic data (discussed ahead).

Also according to the study of Yavuz et al. (2015), the contact VSC/PhylliteQuartzite Group (PQG) provided a moderate reector in drill-hole SZ24A. Deep seismic reectors at a depth of about 2 km (approximately 0.8 s) in prole 1 are also probably related to the VSC/PQG sedimentary rocks contact. The deeper reectors below 2 km may correspond to the geological basement rheo-logic discontinuities. Until the present time, these geological features have not yet been intersected by exploration drill-holes.

Several seismic reectors related to the low-angle thrust planes were found, in both the former CGG prole and prole 1 acquired in this study (see Figs. 5, 6, 7, 8). These reectors have a better lateral continuity in the stacked sections compared to stratigraphy-associated reectors.

1 3

Int J Earth Sci (Geol Rundsch) (2017) 106:899915

909

Fig. 9 Cotovio geological cross-section based on the logs of drill-holes CT01 (Somincor, 1992), CT08001 (AGC/Lundin Mining, 2008) and surface geological data. Reectors observed in the seismic reection prole1 acquired in this study are also shown. See location of drill-hole data in Fig. 11. Numbers indicate total drilled length (not corrected for drill-hole deviation). Geological formations considered: Mrtola Formation (Baixo Alentejo Flysch Group, Upper Visean): shales and greywackes; IPB Volcanic-Sedimentary Complex: sedimentary rocks-Brancanes Fm., shales; Godinho Fm., siliceous shales; Borra-de-Vinho Fm., purple/ green shales; Grandaos Fm., siliceous shales with nodules; Jaspers and cherts; volcanicsrhyolite and rhyodacite

Other holes located close to the seismic proles were used to provide a seismic stratigraphy approach to the seismic interpretation. However, drill-hole data in the study area do not provide information on the stratigraphy at depths over 2 km. Therefore, several deep seismo-stratigraphic packages that could be observed in the seismic proles underneath the PQG remain to be identied, since they were never reached by any drill-holes to date and are not known on the surface. A change of the seismic character (reectivity pattern and frequency content) was identied in all proles below a certain depth, often topped by a moderate to strong reector(s). The seismic horizon that corresponds to pre-Devonian geological units is marked in red in Figs. 5, 6, 7 and 8.

The 3D model building did not include these deep geological units because they are deeper than present-day exploitation capabilities. However, the proles show the presence of tectonic structures, probable low-angle thrusts, that control the geometry of the Palaeozoic basement. These thrusts also show a tangential deformation dominated by a signicant SW vergence. These data are well correlated with the tectonic models dened by the existing exploration drill-hole data, limited to 11.5 km depth.

As stated above, potential-eld data were also used qualitatively to assist in the interpretation, particularly aeromagnetic data which present the largest coverage of the project area. The Bouguer anomaly map was very useful in dening the orientation of Late Variscan vertical strike-slip faults that offset the Variscan geological faults and horizons, controlling Late Variscan horsts and grabens. The Bouguer anomaly map was also used to conrm the geometry of the deeper and higher-density seismo-geological layers. Together with aeromagnetic data, it provided information about depth variations of the VSC along and across the proles, also allowing to infer the presence of possible signicant massive sulphide deposits.

The major characteristics of the seismic proles are the presence of several parallel high-angle thrusts, nappes and the presence of several strike-slip faults with minor normal component that cut across the former thrusts. Those strike-slip faults locally host small Late Variscan Cu (and other metals) vein deposits and occurrences (Carvalho 1991; Matos et al. 2003; Reiser et al. 2011; Matos and Filipe 2013), mostly forming a NWSE Cu lineament anchored in Neves-Corvo mine and in close proximity to Neves-Corvo

1 3

910 Int J Earth Sci (Geol Rundsch) (2017) 106:899915

extended thrust (see location in Fig. 1). A large NNESSW oriented fault, near-parallel and very close to prole 4, probably of Late Variscan age, can be seen on both gravi-metric and magnetic data (Figs. 3, 4).

We have also marked several deep strong seismic reectors within the basement (beneath the pre-Devonian horizon shown in Fig. 5, 6, 7, 8). These deep high-amplitude reectors in vertical incidence reection seismic data are usually associated with either uids or igneous intrusions (e.g. Carbonell et al. 2004; Brown et al. 2012) and increased fracturing (e.g. Schmelzbach et al. 2008a). It is also probable that some of them are out-of-the-plane energy due to a 3D geological environment with intense faulting and strong lateral velocity variations. Those high-amplitude events that cut across other reectors are likely to have this origin.

Some of the deep reectors have a similar orientation, structure and amplitude of those observed in a deep seismic reection prole acquired nearby in Spain, close to the border with Portugal and covering IPB terrains (Schmelzbach et al. 2008a, b). These deep reectors were interpreted as thrusts and fold belts, major faults and sometimes as mac intrusions possibly associated with the VSC (Schmelz-bach et al. 2008b). Similar reectivity patterns observed in another deep seismic reection prole acquired over Palaeozoic terrains of the Central Iberian Zone have also been interpreted as imbricate thrust systems (Ehsan et al. 2014).

The 3D structure of the study area, resulting from the lateral prole-to-prole interpretation, was also constrained using gravimetric and magnetic data. Figure 10 shows a 3D structural model based on the interpretation of the geophysical, drill-hole and geological data. In general, a subsurface 3D model may be obtained by the creation of geological regions within the study area (Almeida et al. 1993; Ferreira et al. 2010; Charifo et al. 2012). In the current research work, the 3D structural models were obtained by the creation of the transition surfaces between the following main groups: ysch (Mrtola and Mira Formations), VSC and PQG. In order to improve the 3D model precision, several geological sections were dened along the study area, introducing a regional geological model. The interpretation of the stacked seismic sections was important, because they delineated the known faults to depth from the surface geological maps and drill-holes, and the identication of formation discontinuities.

The modelling was carried out using gOcad (Paradigm 2014), and all the relevant data were represented in the 3D environment. First, an initial model representing the main geological features was constructed, including the extension of the structural framework from Neves-Corvo to the SE (Spanish border). It was important that the behaviour of structures such as other thrusts, nappes and late faults should be demonstrated as well. Thus, based on geological mapping, drill-hole geological data

and geophysical data, the fault offsets were inferred. Secondly, another 3D model was dened incorporating the fragmentation of the upper and lower basal surfaces of the formations, with the corresponding fault offsets. The resultant model validation was carried out by overlaying drill-hole data and geological mapping with interpreted seismic stacked sections and other geophysical data (Inverno et al. 2015a).

Figure 11 shows the location of the seismic proles over a more detailed geological map after Oliveira et al. (1988). This map also shows the location of a NWSE oriented geological cross-section extracted from the 3D model and therefore based on geological, drill-hole and geophysical data (Fig. 11b). This gure also shows reduced-to-pole total eld magnetic intensity and Bouguer anomaly proles along the cross-section (Fig. 11c).

It is noteworthy that there is an increasing deepening of pre-Devonian basement from seismic prole 1 to prole 4, in particular in the north-eastern part of the proles (Figs. 10a, 11). In proles 5 and 6, particularly in the northeastern side of the proles, the average depth of the pre-Devonian units rises relative to the depths observed in proles 14. At shallower depths, seismic data show that there is a deepening of the ore-bearing VSC unit to the SE from proles 1 to 4. Aeromagnetic anomalies related to the IPB RosrioNeves-Corvo volcanic axis show a loss of amplitude to the SE in this region, which suggests that, taking into account drill-hole, gravimetric and seismic data, there is deepening of the volcanic axis in this area. This deepening is progressive from the Neves-Corvo mine area (see location in Figs. 1 and 11), where the VSC outcrops for a few kilometres, to the SE of it. According to seismic, drill-hole and potential-eld data in the region covered by proles 5 and 6, the top of the VSC is shallower again. Adjacent to the border with Spain, in the Alcoutim region, the VSC does outcrop in a local horst-type structure (Leca 1976).

Discussion

The CRS processing has clearly improved the signal-to-noise ratio of the seismic data, but it also enhanced little side-scattered events and conicting dips (see Fig. 2). This is due to strong geological lateral changes present in the study area and sensitivity of CRS stacking to velocity analysis (e.g. Mann et al. 2007). PSTM removed conicting dips, but at the cost of introducing artefacts. DMO and post-stack migration supplied better results possibly because DMO performs better in anisotropic media (Hale 1984; Anderson and Tsvankin 1997) than the 2D Kirchhoff algorithm used in PSTM which usually cannot handle geological complex media (e.g. Yilmaz 2001).

1 3

Int J Earth Sci (Geol Rundsch) (2017) 106:899915

911

Fig. 10 Two views [a view from SE; b view from WSW]of the 3D structural strati-graphic model constructed after the integration of drill-hole, geological and geophysical data (seismic reection, radiometric, gravimetric and magnetic). VSC Volcanic-Sedimentary Complex, PQG PhylliteQuartzite Group

As evidenced by several authors, to image steeply dipping structures observed in seismic data, DMO and migration are essential processing steps (e.g. Malehmir and Juhlin 2010; Cheraghi et al. 2012). Schmelzbach et al. (2008b) found good results by applying PSTM, preceded by application of an f-k DMO algorithm to seismic data acquired in the IPB close to the study area, which seems to reinforce the importance of DMO followed by an adequate velocity analysis to image correctly geologically complex media.

DMO and post-stack migration did not completely resolve the problem of conicting dips. In part, this was caused by the geological complexity of the media and the presence of out-of-the-plane energy. As also pointed by other authors (e.g. Juhlin et al. 2010; Ehsan et al. 2014), the crookedness of some parts of the seismic lines may have also contributed to this fact. However, if we compare prole 1 (Fig. 5) which had a linear geometry with the other seismic lines (Figs. 6, 7, 8), we are led to think that the complexity of the media was the main cause for the remaining conicting dips after the application of DMO or PSTM.

Anisotropic DMO, advocated for vertically transverse isotropic (VTI) or orthorhombic media (Anderson and Tsvankin 1997; Yilmaz 2001; Tsvankin 2012), was not used in the processing ow due to strong deformation and the steep dips present in the region. These geological aspects, associated with the crooked geometry of the

seismic proles, introduced a strong variation of Thompson parameters along the lines. Therefore, the anisotropy models (e.g. VTI media or the weak anisotropy approximation) under which commercially available anisotropic DMO algorithms are valid are inappropriate for our dataset.

The velocities used to stack the data varied from around 3 km/s at the surface to about 5.5 km/s at 2 s, values that are similar to those used to process other seismic data acquired over the crystalline basement of Iberia (Yavuz et al. 2015; Schmelzbach et al. 2008b; Ehsan et al. 2014; Gonzlez et al. 1998; Flecha et al. 2006). The high-fold data (120), noise suppression techniques such as the use of 6 receivers per station, the high power and frequency content of the seismic source and the long offsets used to record the data all contributed to the good quality of the stacked sections presented here.

On comparing the magnetic signature of the Neves-Corvo mine area and of the region around proles 3 and 4 (Fig. 4), there is a clear loss of magnetic eld intensity around Neves-Corvo in the NWSE to WNWESE oriented anomalies related to the volcanic axis. This is probably due to the presence of more felsic volcanic rocks compared to basic volcanics. There is a clear positive increase in the amplitude of magnetic anomalies related to this axis in the region of the seismic reection prole 6, located SW of Alcoutim, near the Spanish border. The

1 3

912 Int J Earth Sci (Geol Rundsch) (2017) 106:899915

Fig. 11 a Geological mapafter Oliveira et al. (1998) with location of seismic reection proles acquired in this study, b NWSE longitudinal geological cross-section (ab) along the study area based on drill-hole, geological outcrop and seismic reection data and c respective magnetic eld intensity (M) and Bouguer anomaly (Ba) proles. Surface geology.1 seismic reection prole, 2 cross-section, 3 drill-hole and depth of the local geological cross-section, 4 localities, 5 massive sulphide deposits, 6

Cu vein, 7 Mn oxides, 8 BaPb vein, 9 Sb vein

increase in magnetic eld intensity related to the volcanic axis can have three main possible causes: (1) a shallowing of the volcanic rocks; (2) an increase in the volume of volcanic rocks; or (3) an increase of basic volcanic rocks relative to felsic volcanic rocks. Shallowing of the VSC is present since it can be found in outcrop near Alcoutim. Here, slates, siltites, jaspers and volcanic sediments can be found (Leca 1976). About 10 km south-west of Alcoutim, VSC units, including felsic and basic volcanic rocks, have been drilled by Billinton in the 1980s at depths between 0.4 and 0.81 km.

The presented seismic data strongly suggest that the ore-bearing IPB VSC extends from the Neves-Corvo mine region at least to the Spanish border near Alcoutim. They show several thrusts and nappes verging to the south-west and the extension of the tectonic setting of the Neves-Corvo mine area along the volcanic axis (hidden) to the Spanish border and possibly further east (Inverno et al. 2015a). At least three major thrusts that reach the surface or come close to it can be tracked from prole to prole (Figs. 5, 6, 7, 8). On the up-thrust blocks of the faults, potential-eld and drill-hole data show bodies of felsic and basic volcanic

rocks at depths of 400500 m, which are within present-day exploitation capability.

The regions SE of Neves-Corvo (proles 1 and 2) and W of AlcoutimOdeleite (proles 5 and 6) are the most promising in terms of massive sulphide deposits exploration. In the other seismic proles, the geological and geophysical data complexity of the study area precludes for the moment any simple conclusion about the possibility that these volcanic bodies bear massive sulphide deposits. In some parts of the region where these volcanic bodies are placed at shallower depth, a similar geophysical signature to that of the Neves-Corvo mine occurs (i.e. intermediate-low magnetic eld intensity and high Bouguer anomaly). This is the case for the central part of prole 2 (Fig. 6), the centre of the southern part of prole 6 (Fig. 8) and the central part of prole 5 (Inverno et al. 2015a). Other deeper inclined reectors observed in the seismic stacked sections may be associated with thrust planes deeply rooted in the upper crust, similarly to the tectonics apparent in the afore-mentioned nearby ALCUDIA (Martnez-Poyatos et al. 2012; Ehsan et al. 2014) and IBERSEIS (Simancas et al. 2003; Schmelzbach et al. 2008b) proles.

1 3

Int J Earth Sci (Geol Rundsch) (2017) 106:899915

913

Conclusions

Six 2D seismic reection proles were acquired, processed and interpreted to check the possible extension of the Palaeozoic volcanic axis (represented by the Volcanic-Sedimentary Complex) of the Iberian Pyrite Belt beneath the ysch cover to the region SE of the Neves-Corvo mine up to the Spanish border, south of Alcoutim. The geometry of the seismic surveys and the source used were designed to map subsurface features to depths of 45 km. The Unite Sercel wireless equipment allowed exibility and continuous recording across villages and streams, while three synchronized 28-tonne each vibrators provided energy to be recorded down to depths of 10 km.

Several processing sequences were attempted with post-stack and pre-stack time migration, common- and standard common-depth point stacking. Common-reection surface stacking provided a better signal-to-noise ratio, while pre-stack time migration often introduced artefacts due to the highly complex 3D geological environment, with heavy faulting, thrusting and deformation and the lack of good velocity data. The seismic interpretation, carried out in a 3D environment, was complemented with geological, drill-hole, radiometric and potential-eld data. Several SW-verging thrust faults and nappes were identied in the seismic proles. These structures are cut by multiple Late Variscan strike-slip faults, sometimes with minor normal component, also identied by the seismic data.

The VSC units deepen from the region of the Neves-Corvo mine (where they outcrop) and in prole 1 to prole 4, located further SE. East of prole 4, along proles 5 and 6, the VSC rises and is located at depths of 400500 m. Drill-hole data very close to prole 6, located south of Alcoutim and near the Spanish border, show the presence of VSC basic and acid volcanic rocks. The extension of the volcanic axis and of its structural framework from the Neves-Corvo mine to Alcoutim, together with the depth of VSC units in the SE part of the studied area, shows the existence within the IPB of new promising exploration areas. This may lead to the discovery of polymetallic massive sulphide deposits of the Neves-Corvo type.

More detailed geophysical exploration surveys should be carried out in the near future, considering the detailed characterization of geological structures in the volcanic axis(es). The integration of seismic data with other methods (magnetic, gravimetric, drill-holes logging and rock dating) may dene consistent geological/geophysical models and their future investigation with deep exploration drill-holes. In this context, the Neves-Corvo geological sequence becomes a reference guide to the Neves-CorvoAlcoutim IPB region exploration.

Acknowledgments All the facilities in the seismic equipment rental and logistics by Prospectiuni are strongly appreciated by the authors. The authors are grateful to Delm de Carvalho for all his suggestions and encouragements to write this paper. We are also indebted to Musa Manzi, Christopher Juhlin and anonymous reviewers for their constructive criticism which has signicantly improved the manuscript. We are also grateful to LNEG for all support to the logistics of the work, providing equipment, facilities and the personnel involved during the acquisition. Daniel Oliveira, Coordinator of the Department of Mineral Resources and Geophysics of LNEG, is particularly acknowledged. The authors thank all LNEG staff involved in the hard eld work. We are also indebted to Lundin Mining and J. M. Castelo-Branco for providing drill-hole and aeromagnetic data used in this work. The authors thank the European Union for nancing the Promine project (Nano-particle products from new mineral resources in Europe228559) under the scope of the FP7 framework, which allowed the acquisition of the seismic data. We are indebted to Rui Vieira from Mohave Oil for the help provided before acquisition of the seismic reection data.

References

Adam E, Milkereit B, Mareschal M (1998) Seismic reection and borehole geophysical investigations in the Matagami mining camp. Can J Earth Sci 35:686695. doi:http://dx.doi.org/10.1139/e98-022

Web End =10.1139/e98-022

Ahmadi O, Juhlin C, Malehmir A, Munck M (2013) High-resolution

2D seismic imaging and forward modeling of a polymetallic sulde deposit at Garpenberg, central Sweden. Geophysics 78(6):B339B350Almeida J, Soares A, Reynaud R (1993) Modelling the shape of several marble types in a quarry. In: Elbrond J (ed) Proceedings of the XXIV international symposium APCOM, Montreal, vol 3, pp 452459Anderson JE, Tsvankin I (1997) Dip-moveout processing by Fourier transform in anisotropic media. Geophysics 62:12601269 Ayarza P, Juhlin C, Brown D, Beckholmen M, Kimbell G, Pechning

R, Pevzner L, Pevzner R, Ayala C, Bliznetsov M, Glushkov A, Rybalka A (2000) Integrated geological and geophysical studies in the SG4 borehole area, Tagil Volcanic Arc, Middle Urals: location of seismic reectors and source of the reectivity. J Geophys Res 105:2133321352. doi:http://dx.doi.org/10.1029/2000JB900137

Web End =10.1029/200 http://dx.doi.org/10.1029/2000JB900137

Web End =0JB900137 Boogaard MV (1967) Geology of Pomaro region (Southern Portugal). PhD Thesis. Grafsch Centrum Deltro. RotterdamBorn M, Wolf E (1980) Principles of optics, 6th edn. Pergamon,

OxfordBrown D, Zhang X, Palomeras I, Simancas F, Carbonell R, Juhlin

C, Salisbury M (2012) Petrophysical analysis of a mid-crustal reector in the IBERSEIS prole, SW Spain. Tectonophysics 550:3546Calvert AJ, Li Y (1999) Seismic reection imaging over a massive sulde deposit at the Matagami mining camp, Quebec. Geophysics 64:2432. doi:http://dx.doi.org/10.1190/1.1444521

Web End =10.1190/1.1444521 Carbonell R, Simancas F, Juhlin C, Pous J, Prez-Estan A, Gonzlez-

Lodeiro F, Muoz G, Heise W, Ayarza P (2004) Geophysical evidence of a mantle derived intrusion in SW Iberia. Geophys Res Lett 31:L11601. doi:http://dx.doi.org/10.1029/2004GL019684

Web End =10.1029/2004GL019684 Carvalho D (1991) A case history of the Neves Corvo MassiveSulphide deposit and implications for future discoveries, Economic Geology Monograph no 8: historical perspectives of genetic concepts and case histories of famous discoveries, pp 314334 Carvalho J (1995) Study of the South Portuguese Zone and Adjacent

Atlantic margin using geophysical data. M.Sc., Dissertation (in Portuguese), University of Lisbon, Portugal

1 3

914 Int J Earth Sci (Geol Rundsch) (2017) 106:899915

Carvalho D, Conde L, Enrile JH, Oliveira V, Schermerhorn LJGS (1976) Livro-guia das excurses geolgicas na Faixa Piritosa Ibrica, na III Reunio do Sudoeste no Macio Hesprico da Pennsula Ibrica. Comunicaes Servios Geolgicos Portugal 60:271315 (in Portuguese)

Carvalho P, Pacheco N, Beliz A, Ferreira A (1996) ltimos desenvolvimentos em prospeo realizados pela Somincor. Bol. Geologico y Minero ITGE, vol 107 (56), Madrid, 3954Carvalho D, Barriga FJAS, Munh J (1999) Bimodal siliciclastic systems: the case of the Iberian Pyrite Belt. In: Barrie CT, Hannington MD (eds) Volcanic-associated massive sulde deposits: processes and examples in modern and ancient settings, reviews in economic geology 8, pp 375408, Society of Economic GeologistsCarvalho J, Sousa P, Matos J, Pinto C (2011) Ore prospecting in the

Iberian Pyrite Belt using seismic and potential-eld data. J Geophys Eng 8(2):142153Carvalho J, Matias H, Rabeh T, Menezes P, Barbosa V, Dias R, Carrilho F (2012) Connecting onshore structures in the algarve with the southern Portuguese continental margin: the Carcavai Fault Zone. Tectonophysics 570571:151162. doi:http://dx.doi.org/10.1016/j.tecto.2012.08.011

Web End =10.1016/j. http://dx.doi.org/10.1016/j.tecto.2012.08.011

Web End =tecto.2012.08.011 Charifo G, Almeida JA, Ferreira A (2012) Managing borehole samples of unequal lengths to construct a high-resolution mining model of mineral grades zoned by geological units. J Geochem Explor 132:209223Cheraghi S, Malehmir A, Belleeur G (2011) 2-D seismic reection imaging in the Brunswick no. 6 massive sulphide and iron deposits, Bathurst Mining Camp, Canada: implications for crustal architecture and mineral potential. Tectonophysics 506:5572. doi:http://dx.doi.org/10.1016/j.tecto.2011.04.011

Web End =10.1016/j.tecto.2011.04.011 Cheraghi S, Malehmir A, Belleeur G (2012) 3D imaging challenges in steeply dipping mining environment: new lights on acquisition geometry and processing from the Brunswick no. 6 seismic data, Canada. Geophysics 77(5):WC109WC122Cordsen A, Galbraith M, Peirce J, Hardage BA (2000) Planning 3-D land seismic surveys, geophysical developments no. 9, Society of Exploration Geophysicists, TulsaEaton DW (2003) Hardrock seismic exploration. Geophysical Development Series No 10, SEG, Tulsa (USA)

Ehsan AE, Carbonell R, Ayarza P, Mart D, Prez-Estan A, Martnez-

Poyatos DJ, Simancas JF, Azor A, Mansilla L (2014) Crustal deformation styles along the reprocessed deep seismic reection transect of the Central Iberian Zone (Iberian Peninsula). Tectonophysics 621:159174Ferreira A, Charifo G, Almeida J (2010) 3D geologic modeling. The example of the Farim-Saliquinh phosphates mineralization, Geophysical Research Abstracts, vol 12, EGU2010-10927, EGU general assembly conference abstracts, Viena, Austria, 10927 p Flecha I, Carbonell R, Zeyen H, Mart D, Palomeras I, Simancas F,

Prez-Estan A (2006) Imaging granitic plutons along the IBERSEIS prole. Tectonophysics 420:3747Gonzlez A, Cordoba D, Vegas R, Matias LM (1998) Seismic crustal structure in the southwest of the Iberian Peninsula and the Gulf of Cadiz. Tectonophysics 296:317331. doi:http://dx.doi.org/10.1016/S0040-1951(98)00151-6

Web End =10.1016/ http://dx.doi.org/10.1016/S0040-1951(98)00151-6

Web End =S0040-1951(98)00151-6 GPEP (Gabinete para a Pesquisa e Explorao de Petrleos) (1986)

Petroleum Potential of Portugal. GPEP, LisboaHale D (1984) Dip-moveout by Fourier transform. Geophysics

49:741757Heinonen S, Heikkinen PJ, Kousa J, Kukkonen I, Snyder T, David B

(2013) Enhancing hardrock seismic images: reprocessing of high resolution seismic reection data from Vihanti, Finland. J Appl Geophys 93:111. doi:http://dx.doi.org/10.1016/j.jappgeo.2013.03.004

Web End =10.1016/j.jappgeo.2013.03.004 Hubral P (ed) (1999) Macro model independent seismic reection imaging. Journal of Applied Geophysics 42 (3,4)-Special issue

on Karlsruhe Workshop on macro model independent seismic reection imaging. Elsevier, AmsterdamInverno C, Rosa C, Matos J, Carvalho J, Castello-Branco JM, Batista

MJ, Granado I, Oliveira JT, Arajo V, Pereira Z, Represas P, Sol AR, Sousa P (2015a) Modelling of the Neves Corvo area. In: Weihed P (ed) 3D, 4D and predictive modelling of major mineral belts in Europe, chapter 11. Springer, Berlin, pp 231261Inverno C, Diez-Montes A, Rosa C, Garca-Crespo J, Matos J, Garca-

Lobn JL, Carvalho J, Bellido F, Castello-Branco JM, Ayala C, Batista MJ, Rubio F, Granado I, Tornos F, Oliveira JT, Rey C, Arajo V, Snchez-Garca T, Pereira Z, Represas P, Sol AR, Sousa P (2015b) Introduction and geological setting of the Iberian Pyrite Belt. In: Weihed P (ed) 3D, 4D and predictive modelling of major mineral belts in Europe, Chapter 9. Springer, Berlin, pp 191208Juhlin C, Stephens M (2006) Gently dipping fracture zones in Paleoproterozoic metagranite, Sweden: evidence from reection seismic and cored borehole data, and implications for the disposal of nuclear waste. J Geophys Res 111:B09302. doi:http://dx.doi.org/10.1029/2005JB003887

Web End =10.1029/200 http://dx.doi.org/10.1029/2005JB003887

Web End =5JB003887 Juhlin C, Elming S, Mellqvist C, hlander B, Weihed P, Wikstrm A (2002) Crustal reectivity near the ArchaeanProterozoic boundary in northern Sweden and implications for the tectonic evolution of the area. Geophys J Int 150:180197. doi:http://dx.doi.org/10.1046/j.1365-246X.2002.01706.x

Web End =10.1046/j.1365-246X.2002.01706.x Juhlin C, Dehghannejad M, Lund B, Malehmir A, Pratt G (2010)

Reection seismic imaging of the end-glacial Prvie Fault system, northern Sweden. J Appl Geophys 70:307316. doi:http://dx.doi.org/10.1016/j.jappgeo.2009.06.004

Web End =10.1016/j.jappgeo.2009.06.004 Kim JS, Moon WM, Lodha G, Serzu M, Soonavvala N (1994) Imaging of reection seismic energy for mapping shallow fracture zones in crystalline rocks. Geophysics 59:753765Leca X (1976) Le volcano-sdimentaire de la rgion dAlcoutim

(Baixo-Alentejo-Portugal). Comunicaes dos Servios Geolgicos de Portugal, T. LX: 139149Leca X, Ribeiro A, Oliveira JT, Silva JB, Albouy L, Carvalho P,

Merino H (1983) Cadre gologique ds minralisations de Neves Corvo (Baixo-Alentejo, Portugal)Lithostratigraphie, palogographie et tectonique. ditions du BRGM, Orlans, France, Mmoire du BRGM nr. 121-1983Malehmir A, Juhlin C (2010) An investigation of the effects of the choice of stacking velocities on residual statics for hardrock reection seismic processing. J Appl Geophys 72:2838. doi:http://dx.doi.org/10.1016/j.jappgeo.2010.06.008

Web End =10.1016/j.jappgeo.2010.06.008 Malehmir A, Dahlin P, Lundberg E, Juhlin C, Sjstrm H, Hgdahl

K (2011) Reection seismic investigations in the Dannemora area, central Sweden: insights into the geometry of poly-phase deformation zones and magnetite-skarn deposits. J Geophys Res 116:B11307. doi:http://dx.doi.org/10.1029/2011JB008643

Web End =10.1029/2011JB008643 Malehmir A, Durrheim R, Belleeur G, Urosevic M, Juhlin C, White

D, Milkereit B, Campbell G (2012a) Seismic methods in mineral exploration and mine planning: a general overview of past and present case histories and a look into the future. Geophysics 77:WC173WC190Malehmir A, Juhlin C, Wijns C, Urosevic M, Valasti P, Koivisto E

(2012b) 3D reection seismic imaging for open-pit mine planning and deep exploration in the Kevitsa NiCuPGE deposit, northern Finland. Geophysics 77:WC95WC108Mann J, Hubral P, Hocht G, Jaeger R, Mueller T (1999) Applications of the common-reection-surface stack. In: 69th annual international meeting of the society of exploration geophysicists, pp 18291832Mann J, Schleicher J, Hertweck T (2007) CRS stackinga simplied explanation, EAGE 69th annual conference and exhibition extended abstracts, London, B044

1 3

Int J Earth Sci (Geol Rundsch) (2017) 106:899915

915

Manzi MSD, Gibson MAS, Hein KAA, King N, Durrheim RJ (2012)

Application of 3D seismic techniques to evaluate ore resources in the West Wits Line goldeld and portions of the West Rand goldeld, South Africa. Geophysics 77:WC163WC171. doi:http://dx.doi.org/10.1190/GEO2012-0133.1

Web End =10.1190/GEO2012-0133.1 Manzi MSD, Hein KAA, King N, Durrheim RJ (2013) Neoarchaean tectonic history of the Witwatersrand basin and Ventersdorp Supergroup: new constraints from high resolution 3D seismic reection data. Tectonophysics 590:94105. doi:http://dx.doi.org/10.1016/j.tecto.2013.01.014

Web End =10.1016/j. http://dx.doi.org/10.1016/j.tecto.2013.01.014

Web End =tecto.2013.01.014 Martnez-Poyatos D, Carbonell R, Palomeras I, Simancas JF, Ayarza

P, Mart D, Azor A, Jabaloy A, Gonzlez Cuadra P, Tejero R, Martn Parra LM, Matas J, Gonzlez Lodeiro F, Prez-Estan P, Garca Lobn JL, Mansilla L (2012) Imaging the crustal structure of the Central Iberian Zone (Variscan Belt): the ALCUDIA deep seismic reection transect. Tectonics 31:TC3017. doi:http://dx.doi.org/10.1029/2011TC002995

Web End =10.10 http://dx.doi.org/10.1029/2011TC002995

Web End =29/2011TC002995 Matias L (1996) Experimental seismology in crustal structural modeling in Portugal mainland. Ph.D. Dissertation (in Portuguese), University of Lisbon, PortugalMatos JX, Martins L, Rosa C (2003) Parque Mineiro da Cova dos

MourosIGM contribute for the sustainable development of the mining park. IGME, Pub. Museo Geom., no 2, pp 487494 Matos JX, Filipe A, coord. (2013) Carta de Ocorrncias Mineiras do Alentejo e Algarve escala 1:400 000, digital version. LNEG/ATLANTERRA (Eds), Lisboa. ISBN: 978-989-675-029-9, Coauthors: Oliveira D, Inverno C, Rosa C, Batista MJ, Pereira Z, Salgueiro R, Cunha T, Barreira F. http://www.lneg.pt/download/7904/

Web End =http://www.lneg.pt/ http://www.lneg.pt/download/7904/

Web End =download/7904/

Matos JX, Pereira Z, Rosa C, Oliveira JT (2014) High resolution stratigraphy of the PhylliteQuartzite Group in the northwest region of the Iberian Pyrite Belt, Portugal. Comunicaes do LNEG, 7 pp. Actas do Congresso Nacional de Geologia, Porto, Portugal Matte P (1986) Tectonics and plate tectonics model for the Variscan belt of Europe. Tectonophysics 126:329374Menke W (2012) Geophysical data analysis: discrete inverse theory.

Academic Press, AmsterdamOliveira JT (1983) The marine Carboniferous of South Portugal: a stratigraphic and sedimentologic approach. In: Sousa MJL, Oliveira JT (eds) The carboniferous of Portugal, Memoria Servios Geolgicos Portugal 29, Lisbon, pp 337Oliveira JT coord. (1988) Geological map of Portugal scale 1:

200,000 Sheet No 7, coord. Oliveira JT, Servios Geolgicos de Portugal, LisboaOliveira T et al., coord. (1992) Geological Map of Portugal. scale 1:

500,000, Servios Geolgicos de Portugal, LisboaOliveira JT, Carvalho P, Pereira Z, Pacheco N, Korn D (2004) Stratigraphy of the Tectonically imbricated lithological succession of the Neves-Corvo mine region, Iberian Pyrite belt. Implications for the regional basin dynamics. Miner Deposita 34:422436 Oliveira JT, Rosa CJP, Rosa DRN, Pereira Z, Matos JX, Inverno

CMC, Andersen T (2013) Geology of the Neves-Corvo antiform, Iberian Pyrite Belt, Portugal: new insights from physical volcanology, palynostratigraphy and isotope geochronology studies. Miner Deposita 48:749766Paradigm (2014) SKUA-GOCAD - ParadigmR 14.1. Release Notes.

Paradigm, USA

Pereira Z, Matos JX, Fernandes P, Oliveira JT (2008) Palynostratigraphy and systematic palynology of the Devonian and Carbon-iferous successions of the South Portuguese Zone, Portugal. Memria 34, INETI, Lisboa

Radzevicius SJ, Pavlis GL (1999) High-frequency reections in granite? Delineation of the weathering front in granodiorite at Pion Flat, California. Geophysics 64:18281835

Reiser FKM, Rosa DRN, Pinto AM, Carvalho JRS, Matos JX, Guimares F, Alves LC, De Oliveira DPS (2011) Mineralogy and geochemistry of tin- and germanium bearing copper ore from the Barrigo remobilised vein deposit, Iberian Pyrite Belt, Portugal. Int Geol Rev 53:12121238

Relvas JMRS, Barriga FJAS, Ferreira A, Noiva PC, Pacheco N, Bar-riga G (2006) Hydrothermal alteration and mineralization in the Neves-Corvo volcanic-hosted massive sulde deposit, Portugal: I. Geology, mineralogy, and geochemistry. Econ Geol 1014:753790

Sez R, Pascual E, Toscano PM, Almodvar GR (1999) The Iberian type of volcanosedimentary massive sulde deposits. Miner Deposita 34:549570

Schermerhorn LJG (1971) An outline stratigraphy of the Iberian Pyrite

Belt. Boletim Geolgico Mineiro Espaa 82(34):239268 Schmelzbach C, Horstmeyer H, Juhlin C (2007) Shallow 3D seismic reection imaging of fracture zones in crystalline rock. Geophysics 72:B149B160. doi:http://dx.doi.org/10.1190/1.2787336

Web End =10.1190/1.2787336 Schmelzbach C, Zelt CA, Juhlin C, Carbonell R (2008a) P- and S(V)-velocity structure of the South Portuguese Zone fold-and-thrust belt, SW Iberia, from traveltime tomography. Geophys J Int 175(2):689712Schmelzbach C, Simancas JF, Juhlin C, Carbonell R (2008b) Seismic reection imaging over the South Portuguese Zone fold-and-thrust belt, SW Iberia. J Geophys Res 113:B08301. doi:http://dx.doi.org/10.1029/2007JB005341

Web End =10.102 http://dx.doi.org/10.1029/2007JB005341

Web End =9/2007JB005341 Simancas JF, Carbonell R, Gonzlez Lodeiro F, Prez-Estan A, Juhlin C, Kashubin A, Azor A, Martnez-Poyatos D, Almodvar GR, Pascual E, Sez R, Expsito I (2003) The crustal structure of the transpressional Variscan orogen of SW Iberia: the IBERSEIS deep seismic reection prole. Tectonics 22:1062. doi:http://dx.doi.org/10.1029/2002TC001479

Web End =10.1029/ http://dx.doi.org/10.1029/2002TC001479

Web End =2002TC001479 Tsvankin I (2012) Seismic Signatures and analysis of reection data in Anisotropic Media, Geophysical References Series No. 19, Society of Exploration Geophysicists, Third Edition, Tulsa, USA Urocsevic M, Ganesh B, Marcos G (2012) Targeting nickel sulphide deposits from 3D seismic reection data at Kambalda. Australia. Geophysics 77(5):WC123WC132Vermeer GJO (2002) 3-D seismic survey design. Society of Exploration Geophysics, TulsaYavuz S, Kinkela J, Dzunic A, Penney M, Neto R, Arajo V, Ziramov

S, Pevzner R, Urosevic M (2015) Physical property analysis and preserved relative amplitude processed seismic imaging of volcanogenic massive suldesa case study from Neves-Corvo. Port Geophys Prospect 63(4):798812. doi:http://dx.doi.org/10.1111/1365-2478.12269

Web End =10.1111/1365-2478.12269 Yilmaz O (2001) Seismic data analysis: processing, inversion, and interpretation of seismic data, vol 1. Society of Exploration Geo-physicists, Tulsa

1 3