Journal of Applied Geophysics 164 (2019) 117–129

Contents lists available at ScienceDirect

Journal of Applied Geophysics

j ourna l homepage: www.e lsev ie r .com/ locate / j appgeo

Gravity, aeromagnetic and electromagnetic study of the gold and pyritemineralized zones in the Haile Mine area, Kershaw, South Carolina

Saad S. Alarifi a,⁎, James N. Kellogg a,⁎, Elkhedr Ibrahim b,c

a School of Earth, Ocean, and Environment, University of South Carolina, Columbia, SC, 29208, USAb Department of Geology and Geophysics, King Saud University, Riyadh, Saudi Arabiac Department of Geology, Faculty of Science, Mansoura University, Mansoura, Egypt

⁎ Corresponding authors.E-mail addresses: salarifi@geol.sc.edu (S.S. Alarifi), kel

https://doi.org/10.1016/j.jappgeo.2019.03.0110926-9851/© 2019 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 2 October 2018Received in revised form 24 January 2019Accepted 15 March 2019Available online 20 March 2019

New high resolution gravity, electromagnetic (EM) and magnetic data are tested for the detection of dissemi-nated sulfides in the area of the Haile Gold Mine, South Carolina. Geophysical interpretations were constrainedwith densities and mineral concentrations measured for 49,183 samples from 448 drill holes. Positive residualgravity anomalies from spectral analysis correlate with the mineralized ore zone. Similar correlations betweenpositive anomalies and the ore zones are observed in the first vertical derivative of Bouguer anomaly and tilt de-rivatives of the residual gravity. Euler deconvolution of the gravity field also shows numerous shallow sources inthe ore zone. Drill core measurements show that metasediments (Richtex unit), (2.76 g/cm3) and gold bearingsamples (2.73 g/cm3) are slightly denser than metavolcanics (Persimmon Fork Formation), (2.69 g/cm3). Corre-lation coefficients for sample density andmineral percentages are pyrite (0.18) and sericite (0.15), indicating thatpyrite is the main mineral increasing sample density. 2D forward gravity models constrained by dense drillingmatch the predicted depth range of density anomalies fromEuler deconvolution. Drilling results confirm a spatialcorrelation between high densities, high pyrite concentrations, and the mineralized zones. High conductivityanomalies are observed over theHaile ore zone aswell as over themetasediments. Core drilling and 2D inversionshow that high conductivity anomalies coincide with zones of high pyrite concentrations. The magnetic field isdominated by anomalies produced by granite plutons 3 km north and 5 km west of Haile and northwest-trending Jurassic diabase dikes. East-northeast trending linear anomalies have been sampled and dated asAlleghanian lamprophyre dikes providing the first magnetic map of these intrusions in the southeasternUnited States.

© 2019 Elsevier B.V. All rights reserved.

Keywords:Haile gold mineDisseminated sulfidesGravityElectromagneticsMagneticsPyriteAlleghanian lamprophyre dikes

1. Introduction

Gold mines at Haile, Ridgeway, Brewer and Barite Hill are all locatedin the Carolina terrane (Fig. 1), part of a volcanic island arc that formedoff the coast of Gondwana (Hibbard, 2000; Hibbard et al., 2002). Thestudy area (Fig. 2) is located in the northern part of South Carolina be-tween Kershaw and Jefferson. All the gold deposits in the Carolina ter-rane are hosted in similar geologic settings near the contact betweenmetamorphosed volcaniclastic and metamorphosed sedimentaryrocks of Neoproterozoic to Early Cambrian age (Worthington and Kiff,1970). The deposit at Haile consists ofmultiple discontinuous ore bodieswith disseminated gold in silicified and pyrite-rich metasediments(Richtex). The ore bodies trend east-northeast, subparallel to the struc-tural trend of the Carolina terrane (SRK Consulting, 2017).

Many studies have suggested that the depositional environment forthe Neoproterozoic to Early Cambrian volcaniclastic and sedimentary

logg@geol.sc.edu (J.N. Kellogg).

formations reflected a transition from arc-related subaerial depositionto a rift-dominated collapse of the arc below sea level (Feiss, 1982;Feiss and Vance, 1995). Tomkinson (1988) argued that the mineraliza-tion at Haile was not syngenetic butwas controlled by the developmentof shear zones. The gold bearing deposits at Haile were affected by re-gional metamorphism that remobilized and concentrated gold and sul-fides in structurally favorable sites (Speer and Maddry, 1993).

In the Haile Mine area in South Carolina, the ore deposits and bed-rock geology are covered by a 10 to 40 m thick blanket of saprolite,kaolin-rich residuum derived from intense weathering in sub-tropicalclimates. On top of that, an apron of coastal plain sand, the CretaceousMiddendorf Formation (Nystrom Jr. et al., 1991), up to 23 m thick,covers much of the Haile property (SRK Consulting, 2017). As a result,numerous geophysical surveys, including gravity, airborne magnetic,airborne electromagnetic (EM), and induced polarization/resistivity,have been conducted at Haile to define bedrock geology and locatemin-eralized zones. In this paper, we will focus on results from gravity, aswell as present preliminary results from airborne magnetic and EMsurveys.

Fig. 1. Map of the southeastern United States showing major tectonic terranes in this part of the Appalachian mountain chain after Berry et al. (2016). Five gold mines areshown on this map, the Reed mine in North Carolina (NC), and Brewer, Haile, Ridgeway, and Barite Hill, all located in lower metamorphic rank volcanic arc rocks of theCarolina terrane. Alabama (AL); Delaware (DE); Georgia (GA); Indiana (IN); Maryland (MD); North Carolina (NC); Ohio (OH); South Carolina (SC); Tennessee (TN);Virginia (VA); Washington, D.C. (DC), West Virginia (WV).

118 S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

The gravity method is often integrated with magnetic and electro-magnetic methods during integrated base-metal surveys (Telfordet al., 1976). Common ore minerals (magnetite, pyrrhotite and pyrite)have densities above 4.0 g/cm3, so that their presence can increase thebulk density of a rock (e.g., Airo, 2015). High gravity anomalies are re-lated to high density bodies, whereas low gravity anomalies are relatedto low density bodies (Hoover et al., 1995). Gravity surveys can play acritical role in the identification of pyrite ore bodies with associatedpolymetallic massive sulfides, such as in the Iberian pyrite belt ofPortugal (Oliveira et al., 1998). In the Iberian pyrite belt, the ore bodyis hosted by a much less dense volcaniclastic-sedimentary host rock.The large density contrast between the massive sulfides and the hostrocks contributes to a large gravity anomaly. In Spain, the Las Crucesvolcanic massive sulfide deposit, was discovered by a regional gravitysurvey (McIntosh et al., 1999). The deposit occurs at a depth of approx-imately 120 m beneath the Tertiary sediments and is hosted by pyritebelt lithologies that consist of altered felsic to intermediate volcanicrocks with minor sedimentary rocks (McIntosh et al., 1999; Morgan,2012). In New Zealand, the adularia-sericite epithermal gold-silver de-posits in the Waihi district, show a positive Bouguer gravity anomaly.These gold-silver deposits consist mostly of andesite-hosted quartzveins and are characterized by alteration haloes of pervasive clay alter-ation,magnetite destruction, and sulfidemineralization. The epithermaldeposits are associatedwith a striking, enigmatic positive residual grav-ity anomaly (Morrell et al., 2011). Southwest of Waihi, positive gravityresidual anomalies have been identified over the rhyolite-hostedKarangahakedeposit (Harris et al., 2005). In theCanadianYukon, the re-sidual gravity anomalies helped detect the Clear Lake massive sulfidedeposit. The sulfideminerals are laminated and consist of pyrite, galenaand sphalerite (SRKConsulting, 2010). Similar positive residual Bougueranomalies occur over the Vangorda, Faro and Swim Lake deposits in theYukon (Ford et al., 2007).

The magnetic method has much in common with the gravitymethod, but the magnetic field is much more complicated and vari-able. The magnetic field is bipolar, non-vertical in direction, with

sharp local anomaly variations, while the gravity field is unipolar, ver-tical in direction, with smoother and regional anomalies (Telford et al.,1976). Pyrite is non-magnetic and tends to decrease the rock suscep-tibility (Airo, 2015).

Electromagnetic methods have been used quite successfully in thedirect detection of massive and disseminated sulfide mineralization, aswell as for lithological and structural mapping (Dentith and Mudge,2014). Helicopter systems have been effective in near-surface mapping,but depth penetration is limited in areas with conductive overburden(Allard, 2007). Electromagnetic resistivity depends on many factorssuch as porosity, pore conductivity, rock type, the natural fluid and themineral content of the solid matrix (Airo, 2015). Low resistivity is asso-ciated with sulfideminerals, and increased porosity in sericitized rocks,and resistivity highs are associated with silicification and intrusions(Ford et al., 2007).

A helicopter EMand aeromagnetic survey in theDominican Republicdetected anomalies associated with the intermediate sulfidationepithermal style Romero gold deposits with disseminated to semi-massive sulfides (Legault et al., 2016). Low magnetic values over theRomero gold deposits were interpreted to reflect hydrothermal alter-ation. EM resistivity lows (high conductivity) were also located overthe Romero deposits (Legault et al., 2016).

A groundmagnetic and low frequency electromagnetic survey at theHaile mine site in the late 1970s located diabase dikes (Wynn and Luce,1984). Wynn and Luce (1984) indicated that high resistivity responseswith silica content. In the late 1980s, induced polarization (IP) andground magnetic surveys by Piedmont led to the discovery of theSnake ore deposit (Larson andWorthington, 1989). In 2010, an airbornemagnetic survey for Romarco Inc. located the diabase dikes and felsicintrusive plutons, but did not distinguish the older units.

In this paper, we will show that the high resolution residual gravityanomalies correlate with the ore mineralized zones at Haile and offer apossible explanation for this correlation. We also present preliminaryresults for magnetic and EM surveys at Haile, including correlations ob-served between mineralized zones and low resistive anomalies.

Fig. 2.Geological map of the Haile Mine area after SRK Consulting (2017). A) Surface geological map. B) Geological bedrockmap at 300 ft (91m) above sea level. The red polygon outlinesthe southwest area covered by the helicopter survey. The black lines are A and B profiles after Mobley et al. (2014).

119S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

2. Geological setting

The Carolina terrane extends for more than 500 km from Virginia toGeorgia, with a maximum width of 140 km in central North Carolina.The Carolina terrane was accreted during the Paleozoic Acadian-Neoacadian orogenic event (Hibbard, 2000; Hibbard et al., 2002). Anintrusive magmatic and metamorphic overprint is found mostly in thewestern portions of the Carolina terrane as a result of oblique accretionwith Laurentia, progressing from north to south. The Carolina terranecontains low-grade meta-igneous and metasedimentary rocks ofNeoproterozoic to Late Cambrian age (Secor et al., 1983).

In South Carolina, the Carolina terrane consists of a five kilometerthick sequence of metasedimentary rocks of the Emory, Richtex, andAsbill Pond formations. These overlie a three kilometer thick sequenceof metavolcanic rocks, known as the Persimmon Fork Formation(Secor et al., 1983; Hibbard et al., 2002; Secor and Snoke, 2002). ThePersimmon Fork Formation is mainly composed of felsic volcanic

rocks of rhyodacitic to andesitic composition. Themainminerals withinthis unit are quartz, albite, white mica, chlorite, and biotite deposited ina subaerial environment (Snider et al., 2014). The uppermost one kilo-meter of the Persimmon Fork Formation contains clastic metasedi-mentary units that host the ore zones at Haile (SRK Consulting, 2017).The Richtex Formation contains beds of thinmetamorphosed siltstones,mudstones, wackestones, and turbidite deposits. The main mineralswithin this unit are quartz, pyrite (generally less than 10%), mica (upto 50%), pyrrhotite, and chlorite, (Dennis and Wright, 1997; Hibbardet al., 2002; Mobley et al., 2014).

The Persimmon Fork and Richtex formation rocks are dissected bynorthwest-trending Jurassic diabase dikes and are intruded by Carbon-iferous granites. Numerous thin (0.1 to 2m thick) lamprophyric alkalinedikes are also observed in the Haile area (Mauger, 1988; Maddry andKilbey, 1995). The lamprophyre dikes contain biotite and plagioclasewith chlorite and calcite and are mostly east-northeast trending (SRKConsulting, 2017) with subordinate north- and northeast-striking

120 S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

dikes (Hayward, 1992). Ar40/Ar39 weighted plateau ages from biotitesamples yielded early Alleghanian Pennsylvanian ages of approximately311 Ma (Mobley et al., 2014). These ages are similar to the 314 ± 2 Maage of the Dutchman Creek Gabbro that extends 500 km from NorthCarolina to Georgia (Mobley et al., 2014). The magnetic images in thispaper may provide the first opportunity to map the mafic Alleghaniandikes.

The gold deposits at Haile consists of multiple discontinuous orebodies. The gold mineralization is disseminated and associated withpyrite, pyrrhotite and molybdenite. The ore bodies trend northeast-southwest and east-northeast, sub-parallel to the trend of the CarolinaTerrane. Within the mineralized zones, quartz is dominant, pyrite islesser (0.5 to 3%), and sericite is moderate to strong proximal to orezones. Moving away from the mineralized zone, quartz and pyrite de-crease while sericite increases in abundance (Snider et al., 2014; Hulseet al., 2008).

The Haile mine is interpreted as a low sulfide sediment-hosted golddeposit (SRK Consulting, 2017; Nora and Ayuso, 2012). The Re–Os ageof mineralization at the Haile Mine is 548.7 ± 2 Ma, close to the age ofthe host rocks at Haile and Ridgeway, 553± 2 and 556± 2Ma, respec-tively (Mobley et al., 2014). Haile is interpreted to be a hydrothermaldeposit driven by deep-seatedmagmatism. Thus, themineralization oc-curred while the Carolina Terrane was still located in a peri-Gondwanasite (Mobley et al., 2014).

Gold mineralization at Haile is mostly hosted by sheared, laminatedsiltstone and greywacke within 100m of the sediment volcanic contact(SRK Consulting, 2017). Structures include minor shearing and struc-tural disruption as well as isoclinal folding and recumbent folds (Secorand Wagener, 1968; Secor and Snoke, 1978, 2002; Bell III, 1980;Maddry and Kilbey, 1995; Mobley et al., 2014).

Fig. 3. A) Location map of gravity stations. Red triangles - OceanaGold stations), black circC) Helicopter flight lines for magnetic and electromagnetic (EM) surveys. D) Total magnetThe polygon outlines the southwest area covered by the helicopter survey.

3. Data sets and methodology

High resolution ground gravity data (Fig. 3a) were collected overthe Haile area in 2010 by Romarco (now OceanaGold). The high res-olution land gravity data was merged with the regional gravity data(Daniels, 2005; https://pubs.usgs.gov/of/2005/1022/). In addition,49,183 density measurements through 448 drill holes were ob-tained by OceanaGold. An Aeroquest helicopter survey flown forRomarco in 2011 collected high resolution magnetic and timedomain electromagnetic data over an area of 48 km2 in the HaileMine region (Fig. 3c). The survey was flown with a line spacing of50 m over the west block (mine area) and 100 m over the eastblock. The control (tie) lines were flown perpendicular to the surveylines with a spacing of 500 m (west block) and 1000 m (east block).The nominal EM bird terrain clearance was 30 m.

The total magnetic intensity field was reduced to the magnetic pole(RTP) by using the Gx's technique (Phillips, 2007). The RTP was calcu-lated (Fig. 4A) using the inclination and declination values of 63° and−7.20°, respectively. The RTP filter is applied in the Fourier domainand it moves the observed field from the observation inclination anddeclination to what the field would look like at the magnetic pole.Thus, RTP removes the asymmetries caused by a non-vertical magneti-zation (Dobrin and Savit, 1988), where the RTPmagnetic anomalies arelocated directly above the source. The RTP algorithm assumes thatmag-netic responses are magnetized in the direction of the Earth's presentday (normal) magnetic field; if remanent magnetization is present,then magnetic anomalies will not be correctly resolved. The shaded re-lief technique is commonly used to enhance the image (Fig. 4B),highlighting linear structural trends by varying the artificial illumina-tion direction.

les - Open-File (Daniels, 2005). B) Bouguer anomaly map showing all gravity stations.ic intensity map. E) EM conductivity map. Black outline is Haile Mine ore body extent.

Fig. 4. A) Reduced to pole (RTP) magnetic anomalymap. B) Grayscale shaded relief imageof the RTP magnetic anomaly map. Black outline is Haile Mine ore body extent.

121S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

3.1. Isolate shallow short wavelength anomalies (power spectrum)

To separate long (deep) and short (shallow) wavelength potentialfield anomalies, the cutoff wavelengths and information about the con-tribution of the short and long wavelengths in the spectrum can be ob-tained from the calculated radially-averaged power spectrum of thedata using fast Fourier transform (FFT) (Spector and Grant, 1970;Bhattacharya, 1965). Spectral analysis of the potential field anomaliesindicates an ensemble average depth to different sources of anomalies(Reeves, 2005; Rama Rao et al., 2011; Whitenhead and Musselman,2011). The power spectrum of the gravity and magnetic fields for theHaile Mine area can be approximated by two linear segments (Fig. 5aand c). The low frequency segment relates to deeper sources and thehigh frequency segment relates to shallower sources. This methodologyaverages source depths over a region containing complex anomaliesand is less affected by interference due to overlapping anomalies and

high-frequency noise than other methods (Hinze et al., 2013). The fil-tered residual RTP magnetic and Bouguer gravity anomaly maps(Fig. 5b and d)were produced by applying a Butterworth highpass filterto isolate wavelengths less than 0.5 km for RTP and less than 2 kmfor gravity.

3.2. Edge detection (tilt and 1st vertical derivative)

To enhance the signal related to near-surface geological structures,the RTP aeromagnetic and Bouguer gravity data were filtered usingthe first vertical derivative (Fig. 6). Tilt derivative is another methodto detect the edges of shallow geological sources. Fig. 7 shows the tiltderivatives of the filtered RTP and the filtered gravity maps. Tilt deriva-tive (TDR) or tilt angle, or local phase, was first described by Miller andSingh (1994) and refined by Verduzco et al. (2004) and has been devel-oped by Salem et al. (2007, 2008). TDR is a normalized derivative basedon the ratio of the vertical (VDR) and horizontal (THDR) derivatives ofthe field (Salem et al., 2007). The TDR method assumes that the sourcestructures have buried 2D vertical contacts (Salem et al., 2007, 2008).The tilt derivative (Miller and Singh, 1994; Verduzco et al., 2004) is de-fined as TDR = tan−1 (VDR/THDR), where VDR and THDR are first ver-tical and total horizontal derivatives of the total magnetic intensity,respectively. The tilt derivative ranges between ±90o or −π/2 and π/2(radian) regardless of the amplitude of the vertical derivative or the ab-solute value of the total horizontal gradient (Salem et al., 2007, 2008).The zero contour of the tilt derivative map can be used to delineatethe edges of source bodies, and its negative values are outside the source(Miller and Singh, 1994).

3.3. Depth estimation (Euler deconvolution)

We also used the Euler deconvolution (ED) method to estimate thedepth and location of magnetic and gravity sources from the observedfield. The method was developed by Thompson (1982) to interpret a2D magnetic profile and extended by Reid et al. (1990) to be appliedto gridded data. The 3D Euler deconvolution is based on the Euler's ho-mogeneity equation, an equation that relates the potential field (mag-netic or gravity) and its gradient components to the location of thesource, with the degree of homogeneity N, interpreted as a structuralindex. The Euler deconvolution in 3D is given by Reid et al. (1990)

x−x0ð Þ dF=dxþ y−y0ð Þ dF=dyþ z−z0ð Þ dF=dz: ¼ N B−Tð Þ

where (x0, y0, z0) is the position of amagnetic sourcewhose total field Tis observed at (x, y, z). The total field has a regional value of B (back-ground value). N is the structural index (SI). The gradients dF/dx, dF/dy and dF/dz are thefirst derivatives in thedirection of x, y and z respec-tively. The SI parameter value relies on the source body type and the po-tential field (Table 1; Whitehead, 2010; Reid et al., 1990).

An advantage of the Euler deconvolution method is that it is inde-pendent of field direction, dip, or strike of the anomaly feature, so thereduction to pole is unnecessary, as the source positions can be accu-rately reproduced. In addition, this technique assumes no particulargeological model. Euler deconvolution (ED) solutions for the magneticand gravity fields at theHaileMine areawere calculated using structuralindex 0 and uncertainties of 10% and 45% respectively (Fig. 8A and B).The number of grid cells in the x- and y-dimensions, i.e., the windowsizes, are 10 and 20 respectively.

3.4. Location and source of ore zone density anomalies (2D forwardmodelsand drillcore analysis)

To identify the source of the positive gravity residual anomalies overthe Haile ore zone, forward density models were constructedconstrained by the extensive drillhole sampling available. In addition,

Fig. 5. A) Power spectrum of RTPmagnetic map. B) Filtered residual RTPmap. C) Power spectrum of Bouguer anomalymap. D) Filtered residual Bouguer anomalymap. Maps C and D areproduced by applying a Butterworth highpass residual filter with wavelength cutoffs of 0.5 km and 2 km respectively. Black outline is Haile Mine ore body extent.

122 S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

the large drillcore database was analyzed to correlate density withmin-eralogy. 2-D gravity forward models were constructed using OasisMontaj and 2-D GM-SYS software for two north-south geological sec-tions, A and B, across the ore zone (Fig. 2). The models were based onpreviously published geologic profiles (Mobley et al., 2014) andconstrained by core drilling to depths of up to 650 m. Rock densities(Table 2) are from 2016 OceanaGold laboratory measurements of5606 samples from 60 drillcores using the water immersion method.Other mineral densities values (Table 3) were taken from Carmichael(1989) and Telford et al. (1990).

3.5. Conductivity-depth imaging of time-domain electromagnetic (TEM)data

Electromagnetic data over theHailemine area used in this studywascollected with a helicopter-borne time-domain electromagnetic systemwith a concentric horizontal coil configuration. Raw streaming data,sampled at a rate of 36,000 Hz (120 channels, 300 times per second),was processed with Aeroquest Limited software, pre-filtered, stacked,binned, and split into individual line segments. Conductivity-depth

imaging was derived from a pseudolayer half-space model (Palacky,1981; Huang and Rudd, 2008). Effective depth is derived empiricallyfrom the diffusion depth and apparent thickness of the pseudolayer.The pseudolayer technique provides immunity to altimeter errors andbetter identification and resolution of conductive layers than the homo-geneous half-space model.

4. Results

4.1. Positive density anomalies over ore bodies (gravity results)

The Bouguer gravity anomalies in the study area (Fig. 3b) range fromabout −6 mGal to 17 mGal. The low gravity values correlate well withthe Carboniferous Pageland and Liberty Hill granites (Fig. 2b) becauseof the granites lower density (2.63 g/cc) than the surrounding meta-morphic country rock (Table 2). Generally higher gravity anomalies inthe southern part of the study area are found over Richtex metased-iments, and slightly lower anomalies are measured in the central partof the map around the Persimmon Fork metavolcanics (Figs. 2b and3b). The highest Bouguer anomalies northwest of the mine are related

Fig. 6.A) First vertical derivative of RTPmap. B) First vertical derivative of Bouguer gravitymap. Black outline is Haile Mine ore body extent.

Fig. 7. A) Tilt derivative (TDR) of filtered RTP magnetic map (Fig. 5b). B) TDR of filteredBouguer anomaly map (Fig. 5d). Black outline is Haile Mine ore body extent.

Table 1Euler Deconvolution (ED) Structural index (SI) parameter values afterWhitehead (2010).

SI Magnetic field Gravity field

0.0 Contact Sill/dyke/step0.5 Thick step Ribbon1.0 Sill/dyke Pipe2.0 Pipe Sphere3.0 Sphere

123S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

to mafic volcanics. At first, no obvious correlation can be observed be-tween the Bouguer gravityfield and themineralized zone atHaile. How-ever, Euler deconvolution of the Bouguer gravity field (Fig. 8b) shows aremarkable number of source solutions in the Haile ore zone from100 m above sea level to as deep as 500 m below sea level. To thenorth, source solutions dip at low angle to the northeast, produced bythe south edge of the Pageland granite laccolith (Figs. 2b and 8b).

Spectral analysis of the Bouguer gravity field (Fig. 5c) distinguishesthe long wavelength regional and short wavelength residual compo-nents of the field. The regional component reflects a depth range of1.5 to 5.7 kmand an average depth of 4 km (Fig. 5c). The high frequencyresidual components represent a depth range of 10 to 900 m and anaverage depth about 290 m. The entire amplitude range of the filteredresidual gravity anomaly map (Fig. 5d) is less than 0.5 mGal, so a highprecision densely-spaced ground gravity survey was required toachieve this resolution (Fig. 3a). The residual gravity anomaly map

(Fig. 5d) shows a correlation between positive residual anomalies andthe mineralized ore zone in the Haile mine area (also compare to bed-rock geology in Fig. 2b). The lowest residual anomalies (Fig. 5d) arefound over the Persimmon Forkmetavolcanics (Fig. 2b). The tilt deriva-tive (TDR) map of residual gravity (Fig. 7b) shows a similar correlationbetween positive anomalies and the ore zone.

Fig. 8. A) Euler deconvolution (ED) of the TMI magnetic map. Structural Index: 0(contacts). Depth calculated relative to ground surface. B) ED of Bouguer anomaly map.Elevation above sea level. Structural index: 0 (sill, dike, fault).

Table 3Mineral density values after Carmichael (1989) and Telford et al.(1990).

Minerals Density

Pyrite 4.9 g/cc to 5.1 g/ccGold 15.6 g/cc to 19.3 g/ccMolybdenite 4.4 g/cc to 4.99 g/cc

124 S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

The vertical derivatives enhance shallow features and are responsiveto local fluctuations in the gravity field. High frequency anomalies areobserved surrounding the ore bodies in the 1st vertical derivative ofgravity field (Fig. 6b), and the edge detection appears comparable to

Table 2Density values from OceanaGold laboratory measurements of 5606 samples from 60 drillholes in 2016 used for 2D forward models (Fig. 11).

UNIT Count Average density (g/cm3)

Saprolite 49 2.45Visible gold-bearing rock (Au) 9 2.74Metavolcanic (MV) 1916 2.74Metasediments (MS) 3388 2.81Diabase dike 123 2.88

Fig. 9. 2-D forward densitymodeling. Top is profile A. Bottom is profile B. In themiddle is ageological map showing metasedimentary units (green) and the two profile locations.Gravity station spacing is approximately 100 m. Note that the gravity values are filteredresidual anomalies, and the total amplitude of the modelled anomalies is only 0.5 mGal!Geology after Mobley et al. (2014) and SRK Consulting (2017). Polygon densities arefrom Table 3. Density of saprolite: 2.45 g/cc, Coastal Plain sediments: 2.6 g/cc (JamesBerry, personal communication, Romarco). Red areas: ore, blue areas: molybdenite, MS:metasedimentary unit, MV: metavolcanic units, SP: saprolite, CP: coastal plainsediments. For geology map see Fig. 2. 0 m depth is sea level. No vertical exaggeration:scale is 1:1. (For interpretation of the references to colour in this figure legend, thereader is referred to the web version of this article.)

125S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

the highpass filtered residual (Fig. 5d) and slightly better than the TDRanomalies. However, these high frequency vertical derivative anomaliesmay contain artifacts of the denser gravity observation spacing aroundthe ore deposits (Fig. 3a).

4.1.1. Pyrite ore zone density anomalies (forward modeling and drillcoreanalysis)

Two-dimensional forward gravitymodelswere generated along twoprofiles with known geology from multiple drill cores to test whetherthe short wavelength residual gravity could be correlated with rockunits andmineralized zones (Fig. 9), and hencewhether high resolutiongravity could be useful as an exploratory tool. The geology is afterMobley et al. (2014), SRK Consulting (2017), and numerous drill cores(OceanaGold). Unit densities for the 2D forward models are fromOceanaGold laboratory measurements of 5606 samples from 60 drillholes (Table 2). The amplitude of the measured surface residual anom-alies is only 0.5 mGal, and the model is sensitive to themeasured thick-nesses of the saprolite and Coastal Plain sediments. The dense drillcoresampling tightly constrained the geology along the profile with highconfidence and therefore was used as the input for our initial 2D poly-gon model. Because of uncertainties in average rock densities, wewere prepared to vary the polygon densities to fit the observed gravity.Remarkably, we did not have to modify the drill core based polygon ge-ometries or the unit densities to fit the observed residual gravity anom-alies in both profiles, despite the small amplitude of the residual signal(0.5 mGal) and the small density difference between themetavolcanicsand metasediments (0.06–0.07 g/cc). This gives us a high level of confi-dence in the foreward density model. The only requiredmodification tofit the observed gravity was the steeply south-dipping metavolcanic-metasediment contact on the south (right) end of both profiles

Fig. 10. A) Average densities of metasedimentary units (MS), gold bearing metasedimentary u(OceanaGold). Number of measurements: MS: 36,061, AU: 31, MV: 13,091. B) Average pD) Correlation coefficient results for minerals and density: silicification (SILI), sericite (SER),legend, the reader is referred to the web version of this article.)

(Fig. 9) which had not been sampled with the drillcores. The predicteddepths of the source bodies for density anomalies are in agreementwith the depth range of solutions from Euler deconvolution of theBouguer field in the ore zone (Fig. 8b) of 130 m above sea level to500 m below sea level.

Why are themetasediments and gold-bearing rocks denser than themetavolcanics? Densities were measured and mineral concentrationsvisually assessed for 49,183 samples from 448 drill holes in the HaileMine area (Fig. 10) from OceanaGold 2016 drilling as well as from pre-vious drilling by Romarco. The measurements included 36,061metasediments, 13,091 metavolcanics, and 31 gold-bearing samples.The average densities (Fig. 10a) are metasediments: 2.76 g/cm3,metavolcanics: 2.69 g/cm3, and gold bearing samples: 2.73 g/cm3.Fig. 10d shows the correlation coefficients for sample density and min-eral percentages. The highest density correlations are pyrite (0.18) andsericite (0.15), indicating that pyrite is the main mineral increasingsample density. A scatter plot (Fig. 10c) also shows that density and %pyrite have a positive linear correlation: density (g/cc) = 2.73 +0.014 (% pyrite). Fig. 10b shows that the average concentration of pyritein the gold-bearing rocks is 1.5%, in metasediments: 1.1%, andmetavolcanics: only about 0.25%.

Fig. 11 shows drilling results along profile B. The geology afterMobley et al. (2014) (Fig. 11a) is compared to the core sample densities(Fig. 11c) and the laboratory estimated pyrite concentrations (Fig. 12d).The drilling results (Fig. 12) show a spatial correlation between highdensities, high pyrite concentrations, and the mineralized zones.

Airo andMertanen (2008) noted that rock densities tend to increaseas the abundance of sulfides increases, and Airo (2015) pointed out thatthe densities of common ore minerals are all above 4.0 g/cm3, so thattheir presence increases the bulk density of rock. Of the common oreminerals, magnetite, pyrrhotite and pyrite all have densities ~ 5 g/cm3.

nits (AU) and metavolcanic units (MV) from 49,183 measurements from 448 drill holesercentages of pyrite (green) and silicification (blue). C) Sample density vs pyrite %.pyrite (PY), pyrrhotite (PO). (For interpretation of the references to colour in this figure

Fig. 11.A) Geological profile B (location: Figs. 2b and 10) afterMobley et al. (2014). No vertical exaggeration and same scale in all profiles. B) Conductivity inverted for depth along profileB. C) OceanaGold measured core densities from drillholes along profile B. D) OceanaGold lab estimated pyrite concentrations from drillcore along profile B. Small black box in A and B:boundary of Fig. 11c and d. Dashed white line in 11A and 11B: sea level. Black dashed line at 91 m (300 ft) above sea level in 11a is the depth of the bedrock map (Fig. 2b).

126 S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

4.2. Conductivity anomalies over the ore zone (EM results)

Fig. 12 shows the raw conductivity map at the bird altitude of30 m. The map shows high conductivity anomalies over the Haileore zone as well as over the Richtex metasediments (A). Low con-ductivity is observed over the Pageland granite outcrops (B,Fig. 12). In Fig. 11b the conductivity is inverted for depth. Notethat a high conductivity anomaly correlates with the position ofthe main ore zone. Cultural anomalies, such as power lines andwater lines can create artifacts in EM maps. Alarifi (2017) showedthat filtering can reduce the signal from cultural anomalies in theEM data. We do not have conductivity measurements on samples

from the Haile Mine area, so we are unable to conclusively deter-mine the source of the EM anomalies. However, we note fromcore drilling along profile B that high conductivity anomalies coin-cide with zones of high pyrite concentrations (Fig. 11d). Conduc-tion is largely electrolytic, and conductivity depends uponporosity, hydraulic permeability, moisture content, concentrationof dissolved electrolytes, temperature and phase of pore fluid(Airo, 2015). Massive sulfides in general and pyrite have low resis-tivities (Airo, 2015 and Ford et al., 2007) and may contribute to thehigh conductivity in the ore zones. The metasediments also have ahigh pyrite concentration (Fig. 10b), while granites are low in py-rite and generally exhibit high resistivity.

Fig. 12. Conductivity map from EM survey. The black polygons in the center are theore zones from the geological map (Fig. 2b). A: Subcropping Richtex Formationmetasedimentary units (Fig. 2b). B: Pageland granite intrusion.

Fig. 13. RTP magnetic map. A: subcropping extent of the Ritchtex Formationmetasedimentary units (Fig. 2b). B: Pageland granite intrusion.

127S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

4.3. Jurassic diabase and Alleghanian lamprophyre dikes (magnetic results)

Unlike the EM and residual gravity fields, there is no apparentcorrelation between the magnetic field and the ore bearingzones. Total Magnetic Intensity (TMI) anomalies in the map area(Fig. 3d) range from 50,050 to 50,440 nT. The highest value,about 50,444 nT, is located in the northeast part of the maparea over the Pageland granite (felsic intrusion). In general,the Reduced to Pole map (Figs. 4 and 13) shows only minordifferences from the total magnetic intensity map (Fig. 3d).

The 2-dimensional power spectrum of the RTP map (Fig. 5a)shows a distinct high frequency residual component of the mag-netic field with an average depth of 250 m.

The magnetic map (Figs. 3d, 4, and 13) is dominated by anom-alies produced by the Pageland granite intrusion and NW-SEtrending diabase dikes. The Richtex metasediments (A in Fig. 13)are relatively non-magnetic. The shaded relief image of the RTPmagnetic map (Fig. 4b) highlights linear features in the maps. Inparticular, the shaded relief map illuminates the Pageland granitecontact. The relief map (Fig. 4b) also shows the NW-SE trendingdiabase dikes. However, in addition the shaded relief map also illu-minates a series of ENE-WSW trending linear anomalies. One ofthese anomalies appears to spatially correlate with an ENE-WSWtrending lamprophyre dike mapped in the Haile ore zone byHayward (1992). Ar40/Ar39 weighted plateau ages from biotitesamples yielded early Alleghanian Pennsylvanian ages of approxi-mately 311 Ma (Mobley et al., 2014). These ages are similar tothe 314 ± 2 Ma age of the Dutchman Creek Gabbro that extends500 km from North Carolina to Georgia (Mobley et al., 2014). Themagnetic maps presented in this paper may be the first for theAlleghanian age lamprophyre dikes in the southeastern US.The Alleghanian dikes are also visible on the 1st vertical derivativeof RTP map (Fig. 6a) and a little less visible on the tilt derivative(TDR) of residual RTP map (Fig. 7a). The ENE-WSW trendingdikes are also apparent as shallow (0–100 m) magnetic sourcebodies in the Euler Deconvolution map (Fig. 8a).

5. Conclusions

1) Following spectral analysis, the filtered residual Bouguer gravityanomaly map shows a correlation between positive anomalies andthe mineralized ore zone in the Haile mine area. The entire ampli-tude range of the residual gravity anomaly map is less than 0.5mGal, so a high precision densely-spaced ground gravity surveywas required to achieve this resolution. Similar correlations betweenpositive anomalies and the ore zones are observed in the first verti-cal derivatives of Bouguer gravity and tilt derivatives of the filteredresidual gravity.

2) Euler deconvolution of the Bouguer gravity field shows a remarkablenumber of source solutions in the Haile ore zone in the depth rangefrom 130 m above sea level to 500 m below sea level.

3) Densities and mineral concentrations from drillcores show that cor-relation coefficients for sample density and mineral percentages arepyrite (0.18) and sericite (0.15), indicating that pyrite is the mainmineral increasing sample density.

4) 2D forward gravity models constrained by dense drill coring matchthe predicted depth range of density anomalies from Euler deconv-olution. The models are sensitive to the measured thicknesses ofthe saprolite and Coastal Plain sediments. The drilling results con-firm a spatial correlation betweenhighdensities, high pyrite concen-trations, and the mineralized zones.

5) High electromagnetic conductivity anomalies are observed overthe Haile ore zone as well as over the Richtex metasediments.Filtering reduces cultural signals and increases confidence inthe correlation. Core drilling and 2D inversion show that highconductivity anomalies coincide with zones of high pyriteconcentrations.

6) The magnetic field is dominated by anomalies produced by thePageland granite intrusion and NW-SE trending diabase dikes.Richtex metasediments are relatively non-magnetic. The shadedrelief map illuminates a series of ENE-WSW trending linearanomalies that have been dated as Alleghanian age lamprophyredikes.

High precision gravity methods appear to be useful tools for futureexploration for sediment-hosted ore bodies with disseminated gold

128 S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

in silicified and pyrite–rich metasediments in the Carolina terrane.Further sampling and testing will be needed to correlate electromag-netic conductivity anomalies with the mineralogy in the ore zone.

Acknowledgements

We thank OceanaGold and John Jory, Exploration Director, andthe Haile Exploration team for the high resolution gravity, magneticand electromagnetic datasets and drill core data, as well as manyfruitful discussions regarding geological interpretations. Two anony-mous reviewers provided detailed constructive comments. Geosoftgenerously provided academic licenses for potential field modelingsoftware. USGS provided open-file regional gravity and magneticcoverage. The first author is supported by a fellowship from KingSaud University.

References

Airo, M., 2015. Geophysical signatures of mineral deposit types- Synopsis. Geological sur-vey of Finland. Special Paper 58, 9–70.

Airo, M., Mertanen, S., 2008. Magnetic signatures related to orogenic gold mineralization,Central Lapland Greenstone Belt, Finland. J. Appl. Geophys. 64, 14–24.

Alarifi, S.S., 2017. High Resolution Gravity, Helicopter Magnetic, and ElectromagneticStudy, Haile Gold Mine, South Carolina. Master's thesis, Retrieved from. http://scholarcommons.sc.edu/etd/4135.

Allard, M., 2007. On the origin of the HTEM species. In: Milkereit, B. (Ed.), Proceedings ofExploration 07: Fifth Decennial International Conference on Mineral Exploration,pp. 355–374.

Bell III, H., 1980. Preliminary Geologic Map and Section of the Kershaw Quadrangle,Kershaw and Lancaster Counties, South Carolina. U.S. Geological Survey Open File Re-port, pp. 80–226.

Berry, M., Mobley, M., Gillon, A., Yogodzinski, M., Bates, C., 2016. A Neoproterozoicepithermal gold deposit—the Haile Gold Mine, South Carolina, USA. Geol. Soc. Am.Field Guide 42, 1–8.

Bhattacharya, B.K., 1965. Two-dimensional harmonic analysis as a tool for magnetic inter-pretation. Geophysics 30, 829–857.

Carmichael, R.S., 1989. Practical Handbook of Physical Properties of Rocks and Minerals,CRC-Press; 1st Edition.

Daniels, D.L., 2005. South Carolina aeromagnetic and gravity maps and data: a website for distribution of data. Open-File Rep. 2005-1022. https://pubs.usgs.gov/of/2005/1022/.

Dennis, A., Wright, J., 1997. The Carolina terrane in northwestern South Carolina, U.S.A.:late Precambrian-Cambrian deformation and metamorphism in a peri Gondwananoceanic arc. Tectonics 16, 460–473.

Dentith, M., Mudge, S., 2014. Electrical and electromagnetic methods. Geophysics forthe Mineral Exploration Geoscientist. Cambridge University Press, Cambridge,pp. 235–350 https://doi.org/10.1017/CBO9781139024358.007.

Dobrin, M.B., Savit, C.H., 1988. Introduction to Geophysical Prospecting. 4th ed. McGraw-Hill, New York.

Feiss, Geoffrey P., 1982. Geochemistry and tectonic setting of the volcanics of the CarolinaSlate Belt. Econ. Geol. 77, 273–293.

Feiss, P.G., Vance, R.K., 1995. The Carolina terrane gold district: rift-related volcanogenicgold deposits of the southern Appalachian Piedmont [abs]. Geol. Soc. Am. Abstr.Prog. 27, A-27.

Ford, K., Keating, P., Thomas, M.D., 2007. Overview of geophysical signatures asso-ciated with Canadian ore deposits. In: Goodfellow, W.D. (Ed.), Mine-ralDeposits of Canada: A Synthesis of Major Deposit-Types, DistrictMetallogeny, the Evolution of Geological Provinces, and Exploration Methods:Geological Association of Canada, Mineral Deposits Division. Special Publica-tion no vol. 5, pp. 939–970.

Harris, M., Cassidy, J., Locke, C., Mauk, L., Stevens, M., Vidanovich, P., 2005. Geo-physical Characteristics of the Karangahake Epithermal Deposit, Hauraki Gold-field, New Zealand: Ministry of Economic Development, 2005 New ZealandMinerals and Mining Conference, Auckland, New Zealand, Proceedings.pp. 115–124.

Hayward, N., 1992. Controls on syntectonic replacement mineralization in para-sitic antiforms, Haile Gold Mine, Carolina Slate Belt, U. S. A. Econ. Geol. 87,91–112.

Hibbard, J., 2000. Docking Carolina: Mid-Paleozoic accretion in the southern Appala-chians. Geology 28, 127–130 February 2000.

Hibbard, J., Stoddard, E., Secor, D., Dennis, J., 2002. The Carolina Zone: overview ofNeoproterozoic to early Paleozoic peri-Gondwanan terranes along the eastern Flankof the southern Appalachians. Earth Sci. Rev. 57, 299–339.

Hinze, William J., von Frese, Ralph R.B., Saad, Afif H., 2013. Gravity and MagneticExploration, Principles, Practices, and Applications. Press, CambridgeUniversity.

Hoover, D., Klein, D., Campbell, D., 1995. Geophysical Methods in Exploration andMineralEnvironmental Investigations. USGS Open-File Rep., 95–0831. https://pubs.usgs.gov/of/1995/ofr-95-0831/CHAP3.pdf.

Huang, H., Rudd, J., 2008. Conductivity-depth imaging of helicopter-borne TEM databased on a pseudolayer half-space model. Geophysics 73 (3). https://doi.org/10.1190/1.2904984.

Hulse, D., Lane, T., Crowl, W., 2008. NI 43–101 Technical Report on the Haile Mine Project(Lancaster County, South Carolina).

Larson, A., Worthington, J., 1989. Geophysical mapping at the Haile mine. Report for Pied-mont Mining Company (January 1989, 22p).

Legault, Jean M., Jeremy, Niemi, Brett, Jeremy S., Shengkai, Zhao, Zihao, Han, Plastow,Geoffrey C., 2016. Passive airborne EM and ground IP\resistivity results over theRomero intermediate sulphidation epithermal gold deposits, Dominican Republic.Explor. Geophys. 47 (3), 191–200.

Maddry, J.W., Kilbey, T.R., 1995. Geology of the Haile gold mine, selected mineral de-posits of the Gulf Coast and southeastern United States: Part II. Gold Deposits ofthe Carolina Slate Belt: Society of Economic Geologists Guidebook Series. vol.24, pp. 147–172.

Mauger, R.L., 1988. Geochemical evidence for sediment recycling from North Carolina(U.S.A.) minettes. Can. Miner. 26, 133–141.

McIntosh, S.M., Gill, J.P., Mountford, A.J., 1999. The geophysical response of the Las Crucesmassive sulphide deposit. Explor. Geophys. 30 (3–4), 123–133.

Miller, H.G., Singh, V., 1994. Potential field tilt – a new concept for location of potentialfield sources. J. Appl. Geophys. 32, 213–217.

Mobley, R., Yogodzinski, G., Creaser, R., Berry, J., 2014. Geologic history and timing ofmineralization at the Haile gold mine, South Carolina. Econ. Geol. 109,1863–1881.

Morgan, L., 2012. Geophysical characteristics of volcanogenic massive sulfide deposits.Scientific Investigations Report 2010–5070–C. U.S. Geological Survey, Reston, Vir-ginia, pp. 115–131.

Morrell, A., Locke, C., Cassidy, J., Mauk, J., 2011. Geophysical characteristics of Adularia-sericite epithermal gold-silver deposits in the Waihi-Waitekauri region, NewZealand. Econ. Geol. 106, 1031–1041.

Nora, K., Ayuso, R., 2012. Gold Deposits of the Carolina Slate Belt, Southeastern UnitedStates: Age and Origin of the Major Gold Producers. USGS, open-file report 2012.p. 1179.

Nystrom Jr., P.G., Willoughby, R.H., Price, L.K., 1991. Cretaceous and Tertiary stratigraphyof the upper Coastal Plain, South Carolina. In: Horton Jr., J.W., Zullo, V.A. (Eds.), TheGeology of the Carolinas: Knoxville. The University of Tennessee Press, Tennessee,pp. 221–240.

Oliveira, V., Matos, J., Bengala, M., Silva, N., Sousa, P., Torres, L., 1998. Geology andgeophysics as successful tools in the discovery of the Lagoa Salgada Orebody(Sado Tertiary Basin–Iberian Pyrite Belt), Grandola, Portugal. Miner. Deposit.33, 170–187.

Palacky, G.J., 1981. The airborne electromagnetic method as a tool of geological mapping.Geophys. Prospect. 29, 60–88.

Phillips, J.D., 2007. Geosoft eXecutables (GX's) developed by the U.S. Geological Survey,Version 2.0, with notes on GX development from Fortran Code. U.S. Geological Sur-vey. Open File, Report, p. 1335.

Rama Rao, C., Kishore, R.K., Pradeep Kumar, V., Butchi Babu, B., 2011. Delineationof intra crustal horizon in Eastern Dharwar Craton – an aeromagnetic evi-dence. J. Asian Earth Sci. 40 (2), 534–541. https://doi.org/10.1016/j.jseaes.2010.10.006.

Reeves, C., 2005. E-Book on Aeromagnetic surveys: Principles, Practice and Interpretation.Retrieved March 2, 2017, from. http://www.geosoft.com/media/uploads/resources/technical-papers/Aeromagnetic_Survey_Reeves.pdf.

Reid, A.B., Allsop, J.M., Granser, H., Millett, A.J., Somerton, I.W., 1990. Magnetic in-terpretation in three dimensions using Euler deconvolution. Geophysics 55,80–91.

Salem, A., Williams, S., Fairhead, D., Ravat, D., Smith, R., 2007. Tilt-depth method: a simpledepth estimation method using first-order magnetic derivatives. Lead. Edge 26,1502–1505.

Salem, A., Williams, S., Fairhead, J.D., Smith, R., Ravat, D.J., 2008. Interpretation ofmagneticdata using tilt-angle derivatives. Geophysics 73, L1–L10.

Secor, T., Snoke,W., 1978. Stratigraphy, structure, and plutonism in the Central South Car-olina Piedmont. In: Snoke, A.W. (Ed.), Geological Investigations of the Eastern Pied-mont, Southern Appalachians: Carolina Geological Society Guidebook for 1978Annual Meeting, pp. 43–63.

Secor, T., Snoke,W., 2002. Geologic map of the Batesburg and Emory Quadrangles, Lexing-ton and Saluda counties, South Carolina. Boulder, Colorado, Geological Society ofAmerica Map and Chart Series MCH. 091, p. 32.

Secor, T., Wagener, D., 1968. Stratigraphy, structure and petrology of the Piedmont inCentral South Carolina. Columbia, South Carolina, South Carolina Geological Survey,Geologic Notes. vol. 12, pp. 67–84.

Secor, T., Samson, L., Snoke, W., Palmer, R., 1983. Confirmation of the Carolina Slate Belt asan Exotic Terrane. Science 221, 649–651.

Snider, J., Patterson, E., Gochnour, L., Marek, J., Burkhalter, C., 2014. Haile Gold Mine Pro-ject, NI 43–101 Technical Report.

Spector, A., Grant, F.S., 1970. Statistical models for interpretation of aeromagnetic data.Geophysics 35, 293–302.

Speer, W.E., Maddry, J.W., 1993. Geology and recent discoveries at the Haile goldmine. Lancaster County, South Carolina: South Carolina Geology. vol. 35,pp. 9–26.

SRK Consulting, Arseneau, G., MacIntyre, D., 2010. Clear Lake Zinc-Lead-Silver DepositYukon. Project Number 2CC035.000.

SRK Consulting, Carr, D., van Brunt, B., Jory, J., Howw, P., Poeck, J., Osbom, J., Moore, J.,Tinucci, J., Swanson, B., Kinakin, D., Malensek, G., Bird, D., Stryhas, B., Prosser, B.,2017. NI 43-101 Technical Report: Feasibility Study, Haile Gold Mine, LancasterCounty, South Carolina (319 pp).

129S.S. Alarifi et al. / Journal of Applied Geophysics 164 (2019) 117–129

Telford, W.M., Geldart, L.P., Sheriff, R.E., Keys, D.A., 1976. Applied Geophysics. CambridgeUniversity Press.

Telford, W.M., Geldart, L.P., Sheriff, R.E., 1990. Applied Geophysics Second Edition. Cam-bridge University Press, New York (744 p).

Thompson, D.T., 1982. EULDPH; a new technique for making computer-assisted depth es-timates from magnetic data. Geophysics 47 (1), 32–38.

Tomkinson, M.J., 1988. Gold mineralization in phyllonites at the Haile mine, South Caro-lina. Econ. Geol. 83, 1392–1400.

Verduzco, B., Fairhead, J.D., Green, C.M., MacKenzie, C., 2004. New insights into magneticderivatives for structural mapping. Lead. Edge 23, 116–119.

Whitehead, N., 2010. montaj Grav/Mag Interpretation, Tutorial and User Guide. GeosoftIncorporated.

Whitenhead, N., Musselman, C., 2011. Montaj MAGMAP Filtering. Geosoft Incorporated.Worthington, J.E., Kiff, I.T., 1970. A suggested volcanogenic origin for certain gold deposits

in the slate belt of the Carolina piedmont. Econ. Geol. 65, 529–537.Wynn, J., Luce, R., 1984. Geophysical methods asmapping tools in a strata-bound gold de-

posit; Haile Mine, South Carolina slate belt. Econ. Geol. 79 (2), 382–388.