Geodynamic constraints on orebody localization in the Anqing orefield, China: Computational modeling and facilitating predictive exploration of deep deposits

Ore Geology Reviews 43 (2011) 249–263

Contents lists available at SciVerse ScienceDirect

Ore Geology Reviews journal homepage: www.elsevier.com/locate/oregeorev

Geodynamic constraints on orebody localization in the Anqing orefield, China: Computational modeling and facilitating predictive exploration of deep deposits Liangming Liu ⁎, Changlin Wan, Chongbin Zhao, Yilai Zhao Computational Geosciences Research Centre, Central South University, Changsha 410083, China

a r t i c l e

i n f o

Article history: Received 10 February 2010 Received in revised form 3 September 2011 Accepted 4 September 2011 Available online 8 September 2011 Keywords: Skarn deposit Mineral system Metallogeny Geodynamics Computational modeling Exploration targeting Anqing orefield

a b s t r a c t The Anqing Fe–Cu skarn deposit, with an age of 134 to 142 Ma and resources of 62.4 Mt at 0.906% Cu and 32.2% Fe, is one of the most important deposits in the Yangtze River Metallogenic Belt, East China. To better understand the localization of orebodies and thus facilitate predictive exploration of deep orebodies, computational modeling is used to simulate the coupled geodynamic processes during the syn-tectonic cooling of the ore-related intrusion, based on geological and geophysical investigations in the Anqing orefield. The occurrences of the ore veins and veinlets in diorite and skarn, as well as the sharp zigzag boundary of the orebody, indicate that the Cu ores were deposited after the solidification of the diorite and skarn formation, and were located in some tensional structural spaces that are unevenly distributed along the contact zone between the felsic intrusion and sedimentary carbonates. The locations of orebodies are closely associated with the contact zone shape. The computational results of two models with two typical contact-shapes show that pore fluid flow was focused into the dilation zones from different sources. All the significant dilation zones, in which the existing orebodies were located, are distributed in some specific places of the south contact zone of the intrusion. In addition, these dilation zones are closely related to the contact zone shape of the intrusion and can control the location of orebodies through the coupled mechano-thermo-hydrological processes during cooling of the intrusion in the extension setting. The skarns are not critical for controlling the localization of orebodies. This means that exploration for deep ore should target deep dilation zones close to the contact boundary of the intrusion. Such recognition may provide a useful guide in selecting exploration targets in the Anqing orefield. As a direct result of computational modeling, an orebody has been discovered in the deep dilation zone in this orefield. It demonstrates that computational modeling is a promising tool for understanding the metallogenic processes and for facilitating the deep exploration of hidden orebodies that are related to intrusions. © 2011 Elsevier B.V. All rights reserved.

1. Introduction Metallogenic studies have made some impressive contributions to the explanation of the general formation and processes involved in ore deposits (e.g., Candela et al., 2005; Franklin et al., 2005; Kerrich et al., 2005; Meinert et al., 2005). Traditional metallogenic theories are, however, not always suitable for conducting sound exploration strategy. The reason for this is the fact that it can often be difficult to obtain sufficient detail concerning an ore-forming system and the critical geodynamic processes responsible for formation and localization of orebodies (Etheridge and Henley, 1997; Liu, 2007; Liu and Peng, 2005; Liu et al., 2005a,b; Price and Stoker, 2002). From this point of view, mineral exploration is still an activity with high economic risk (e.g., Kreuzer et al., 2008; Liu et al., 2005b). A strategy to reduce exploration risk is to apply new technologies for effective target selection and orebody

⁎ Corresponding author. Tel.: + 86 731 88660096. E-mail address: [email protected] (L. Liu). 0169-1368/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.oregeorev.2011.09.005

detection. The generation of targets involves the use of an integrative methodology that comprises metallogenetic concepts, geophysics, spatial analysis, mineral economics, decision making and probability theory (Hronsky and Groves, 2008). Since the appropriate generation of targets has the greatest potential of finding undiscovered orebodies, some new techniques such as fuzzy-logic (Knox-Robinson, 2000; Porwal et al., 2003), mineral systems approach (Knox-Robinson and Wyborn, 1997; Kreuzer et al., 2008; Partington, 2010; Wyborn et al., 1994), fractal analysis (Carranza and Sadeghi, 2010; Ford and McCuaig, 2010), and geodynamic modeling (Hobbs et al., 2000; Liu et al., 2005a; Mair et al., 2000), have been applied in the process of generating targets. The formation of metallic deposits should involve sources that supply metal, fluid and ligand components, and a geodynamic system for transporting the components and depositing the metallic commodity. The geodynamic system is definitely more responsible for the localization of orebodies. Generally, a geodynamic system involves the full feedback coupling between the following five processes: mechanical deformation, pore fluid flow, heat transfer, mass

250

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

transport and chemical reactions (Zhao et al., 1999). The formation and localization of an orebody in a specific place are due to the fact that the related feedback coupling mechanisms reinforce one another in such a way that the rate of mineralization, integrated over time, is optimized in the ‘specific place’ (Hobbs et al., 2000). This kind of feedback coupling mechanism results in the complexity of the mineralization system that is usually beyond the capability of the traditional research methods (Liu, 2007; Liu et al., 2005a). Because the complex system has two fundamental characteristics: an emergence and an attractor (Zhao et al., 2008), a slight change in initial conditions would result in entirely different outcomes. The advancement of computer technology and computational algorithms has made computational simulation an indispensable method for solving complex problems such as those described by the geodynamic systems of coupled metallogenic processes (Hobbs et al., 2000; Liu et al., 2005a,b, 2008; President's Information Technology Advisory Committee, 2005; Zhao et al., 2006, 2007, 2009, 2010). During the last two decades, computational modeling has been widely used to understand the geodynamic processes that lead to the formation of ore deposits (Chi and Savard, 1998; Garven and Freeze, 1984; Liu and Zhang, 2007; Oliver et al., 2001; Ord et al., 2002; Raffensperger and Garven, 1995; Schaubs and Zhao, 2002; Sorjonen-Ward et al., 2002; Yang et al., 2006; Zhang et al., 2007). The application of such modeling to brown field has great potential for enhancing the efficiency of exploration and facilitating the predictive discovery of hidden orebodies. The Anqing orefield is a famous Fe–Cu brownfield ore district in China, and the Anqing deposit is the largest single Fe–Cu skarn deposit in the Yangtze River Metallogenic Belt (YRMB) (Chang et al., 1991; Liu et al., 2008). Because of both economic and scientific importance, the Anqing orefield has been drilled extensively by the Anhui 326 geological team (1976, 1988, 1992), and investigated intensively by many researchers (e.g., Chang et al., 1991; Dong and Qiu, 1993; Liu et al., 2008; Mao et al., 2006; Wang and Zhou, 1995; Zhang et al., 2008; Zhou et al., 2007). This has resulted in the reserve estimation of discovered orebodies and an understanding of ore formation processes. However, as the shallow orebodies have been exhausted by mining, exploration has to inevitably target blind orebodies located at depth. Because of a lack of detectable geophysical or geochemical anomalies, the existing knowledge models and methods cannot predict such blind orebodies. Therefore, it is necessary to develop innovative knowledge models for understanding ore-forming processes in detail, so as to facilitate the predictive discovery of deep orebodies. The geodynamic system that controls the scale and location of orebodies in the Anqing orefield is very complex. It involves coupled multi-processes and multi-factors rather than a simple geodynamic process or a simple linear combination of several factors. For this reason, computational modeling may become a practicable method for solving such a complex geodynamic problem. In this paper, we construct geodynamic models capturing coupled MTH (mechano-thermo-hydrological) processes, based on detailed geological and geophysical investigations in the Anqing orefield. We then carry out computational modeling experiments of these models. The computational results are used to investigate the ore genesis and to predict the location of orebodies so as to select the appropriate targets and therefore lower the exploration risk. 2. The Anqing orefield 2.1. Geological setting The Anqing orefield consists of ore deposits originally associated with the Yueshan intrusion in Anqing, Anhui province, eastern China. It is a major deposit in the Tongling–Anqing District (TAD), which is located in the central segment of the Yangtze River metallogenic belt along the northern margin of the Yangtze craton (Yangtze plate) (Fig. 1). The district is bordered by the Dabieshan UHP (ultra-

high pressure) metamorphic belt and North China craton (North China plate) to the north (Liu and Peng, 2003; Pan and Dong, 1999). The collision of the Yangtze plate with the North China plate took place in the late Triassic. This tectonic event reactivated the Yangtze River fracture zone and produced extensive intermediate to felsic magmatism with the related mineral deposits (Chang et al., 1991; Pan and Dong, 1999; Zhai et al., 1996). The Tongling–Anqing district is a major supplier to the Chinese copper industry. Six major Cu and Fe–Cu deposits have been discovered in this district to date (Fig. 1). They are the Fenghuangshan Cu deposit (the ore reserve of 32 Mt at 1.26% Cu), Shizhishan Cu deposit (the ore reserve of 61 Mt at 1.03% Cu), Dongguashan Cu deposit (the ore reserve of 98 Mt at 1.01 Cu), Tongguangshan Cu deposit (the ore reserve of 36 Mt at 1.16% Cu), Tongshan Cu deposit (the ore reserve of 23 Mt at 1.11% Cu), and the Anqing Fe–Cu deposit (the ore reserve of 62.4 Mt at 0.906% Cu and 32.2% Fe). These deposits are all genetically and spatially associated with the felsic intrusions of a late Jurassic age (Fig. 1) (Liu et al., 2008). Rocks in the Tongling–Anqing district include Precambrian metamorphic rocks and Paleozoic through Mesozoic sedimentary rocks. The mid-Carboniferous to mid-Tertiary rocks are littoral to neritic carbonates interbedded with bathyal facies beds, alternating with marine-continental clastics. They are the most favorable wall rocks to host Cu- and Fe–Cu-skarn deposits (Figs. 2 and 3). The Yueshan diorite intrusion forms the core of the Anqing orefield (Fig. 3). All Cu and Fe orebodies occur strictly in the vicinity of the contact zone between the Yueshan intrusion and the Triassic sedimentary carbonates. The Yueshan diorite is mainly composed of 67.9% plagioclase, 6.1% K-feldspar, 2.2% quartz, 18.1% hornblende and 1% biotite (Anhui 326 Geological Team, 1992). This intrusion is also the largest copper mineralizing intrusion in the district with a surface area of about 3 km 2 (Figs. 1, 3 and 4). The exposed part of the intrusion is uniquely cross-shaped (Fig. 3a). In a 3D view, the distinct feature of the intrusion is that the southern boundary is much more complex in shape than the northern, but the 3D outline of the intrusion (Fig. 4) does not look like a cross as the 2D outline in a surface plane (Fig. 3a). The SHRIMP U–Pb zircon age of the Yueshan intrusion is 138.7 ± 0.5 Ma (Zhang et al., 2008). The Re–Os age of molybdenite in the ores is 134.7 to 142.6 Ma (Mao et al., 2006). The Yueshan intrusion and associated mineralization were emplaced in the late Jurassic to early Cretaceous, when the crustal regime in the YRMB changed from compression to extension (Dong and Qiu, 1993; Mao et al., 2006; Zhang et al., 2008). In the Anqing orefield, there are five main structures: (1) a NWtrending fold with parallel faults along the limbs in the southwestern sector of the field; (2) NS-trending folds with almost parallel faults along the limbs in the southeastern sector of the field; (3) NE-trending folds with parallel faults along the limbs in the northeastern sector of the field; (4) approximately EW-trending faults in the center of the field and (5) NNW- and NNE-trending faults scattered in almost the whole field (Fig. 3). The NW- and NS-trending folds might have been formed immediately after the early Triassic, because the youngest folded strata are of an early Triassic age. The NE-trending folds might have formed immediately after the middle Jurassic, because the youngest folded beds are of a middle Jurassic age. The approximate EW-trending faults consist mainly of normal faults and cut through the NE-trending fold and the Yueshan intrusion. This indicates that there was an extensional event after the Jurassic folding and the magmatic intrusion. The NNW- and NNE-trending faults are certainly the latest, as they cut through all other structures and orebodies. A major approximately EW-trending normal fault is parallel to and immediately nearby the orebodies (Fig. 3). Major orebodies are all located in the approximately EW-trending tensile fractures, while orebody No. 3 is an EW-trending vein in the intrusion. This implies that the extensional event after the Jurassic folding could be related to mineralization.

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

251

Fig. 1. Geological map of the Tongling-Anqing district, showing the location of the Anqing deposit. Modified from Anhui Provincial Geological Bureau (1987).

All ore deposits discovered in the Anqing orefield have the following two types: Cu-bearing quartz veins in the diorite and Fe–Cu skarn in the contact zone. Because the Cu-bearing quartz veins are very small-scale and account for b2% of the total Cu reserves, our research is focused on the skarn orebodies, most of which are distributed in the Anqing ore deposit (Anqing mine). All the Fe–Cu skarn orebodies are located in the vicinity of the contact zone between the Yueshan intrusion and the host rocks, which are composed of: (1) dolomite and limestone breccia at the low part of the Yueshan Formation of a middle Triassic age (T2y 1), and (2) limestone of the Nanlinghu Formation of a lower Triassic age (T1n) (Figs. 2 and 3). The major ore minerals in the skarn ores are magnetite, chalcopyrite, pyrite and pyrrhotite. The minor minerals are hematite, marcasite, bornite, molybdenite, galena and sphalerite. The major gangue minerals are garnet, diopside, with lesser amounts of calcite, actinolite–tremolite, chlorite, quartz, talc, phlogopite, scapolite and serpentine. 2.2. Distribution of orebodies In the Anqing orefield, more than 98% of the discovered ores are of Fe–Cu skarn type, including Cu-bearing skarn and Cu-bearing magnetite ore. These ores are located in the contact zone between the Yueshan intrusion and the host rocks consisting of lower-middle Triassic carbonates. The distinct feature of the orebody localization is displayed by the irregular distribution of orebodies along the contact zone (Liu et al., 2008). More than 90% of the Fe–Cu reserves measured in the Anqing orefield are from the Anqing Fe–Cu deposit, which is located along the east branch of the Yueshan intrusion. In the Anqing deposit, economic orebodies are all discovered in the southern contact zone of the eastern branch of the Yueshan intrusion (Figs. 3, 5 and 6). We think such uneven distribution of orebodies could be attributed to the morphological variation of the contact zone rather than the chemical variation in the host rocks. This interpretation is based on the following association of orebodies with their locations.

(1) Although the major orebodies are distributed along the contact zone between diorite and the banded marble and dolomitic marble, rather than that of quartz diorite with hornfels, the several parts of the contact zone consisting of diorite with banded marble and dolomitic marble are also barren of mineralization. This indicates that the petrologic composition of the wall rocks of the contact zone does not play a controlling role in the localization of orebodies. (2) Although mineralization is associated with the skarn alteration, the scale of orebody is not in proportion to the scale of the skarn in the host rocks. Some orebodies, such as orebody No. 2, are not even within or near the skarn (Liu et al., 2008). This indicates that the skarn does not control the location of orebodies. (3) The common feature of the mineralized contact zones is displayed by the contact zone shapes. The mineralized contact zones are more complex in morphology than the barren contact zones (Fig. 5). In the Anqing deposit, the mineralized segments only occupy a small part of the whole contact zone, but exhibit a regular variation in space and shape. From east to west, the “step-shaped” contact zone becomes a “tongueshaped” contact zone. In the former, the contact zone of the intrusion with the carbonate hanging wall changes from a steep-dip into a gentle-dip; while in the latter, the marble is surrounded by the diorite. The orebodies are only located in the steep segments of the “step-shaped” contact zone and the tips of the “tongue-shaped” contact zone (Fig. 5). This type of unevenly spatial distribution of orebodies along the contact zone needs to be interpreted by a geodynamic model that describes the mineralization processes associated with the Yueshan intrusion. In particular, the geodynamic system to be considered should include the role of the contact zone played in mineralization and orebody localization.

252

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

Orogeny (Time) Yanshanian Orogeny (135Ma)

Epoch Upper Cretaceous Middle Jurassic

Lower Jurassic Indo-Sinian Orogeny (195Ma) Upper Triassic

Middle Triassic

Lithostratigraphic Unit Code Thickness and Lithologic Composition

Favor for hosting Cu skarn

>3000m, conglomerate and gravel-bearing Xuannan Formation K2x sandstone 2168m, feldspar quartz sandstone, siltstone and silty shale, conglomerate at bottom 624m, feldspar sandstone, interlayers of Moshan Formation J1m silty shale and coal, conglomerate at bottom Lalijian Formation T3l 17~35 m, fine sandstone and siltstone with interlayers of carbonaceous shale and coal 1735m, siltstone, interlayers of sandy Tongtoujian T t 2 Formation shale and interlens of fine sandstone 43m, siltstone, interlayered lens of marl in Upper T 2y2 lower part Yueshan Formation 1 Low T 110m, dolomite and limestone breccia 2y Luoling Formation

J2l

>420m, thin- to mid-thick limestone, with Nanlinghu Formation T1n argillic bands in upper and lower Lower Triassic Helongshan Foration T1 h 250Ma

Lower Permian 290Ma

439Ma

Upper Carboniferous Middle Carboniferous Upper Devonian

Gufeng Formation

P1g 35~41m, siliceous slate and siliceous shale

218~246m, chert-bearing bioclastic limestone and carbonaceous shale in lower 25m, orbicular limestone and bioclastic Chuanshan C c 3 Formation limestone 74~141m, bioclastic limestone, dolomite Huanglong C h 2 Formation with fine quartz conglomerate

Qixia Formation

P1q

Wutong Formation D3w 29~94m, quartz sandstone and silty shale

Mid Silurian

Fentou Formation

Lower Silurian

Gaojiabian Formation

Upper Ordovician

Caledonian Orogeny (600Ma)

42m, argillic band-bearing limestone

60m, limestone with interlayers of silt Yingkeng Formation T1y shale 28~58m, siliceous shale and argilloDalong Formation P2d calcareous shale, interlayers of limestone Upper Permian 58~73m, fine sandstone, silt shale and Longtan Foration P2l shale, interlayers of coal

S2f 115m, sandstone, siltstone and sandy shale S1g

Wufeng Formation O3t

776m, graptolite-bearing black shale 2.8~5m, black siliceous shale

2.5~6.4m, calcareous shale with Tangtou Formation O3t interlayers of limestone 21~49m, limestone, interlayers of thin Middle Tangshan Formation t O 2 Ordovician slate 194m, limestone in upper and dolomite in Lower Lunshan Formation O1l Ordovician lower Huangjiabang Cambrian C 175m, limestone siliceous inter-cores Formation

Precambrian

Dongling Group Pt3d >1500m, biotite quartz schist and gneiss

Fig. 2. Stratigraphy and orogenic events in the Tongling–Anqing district, showing that some lithostratigraphic units are favorable for Cu skarn mineralization according to their hosting copper reserves. Data from Anhui Provincial Geological Bureau (1987).

2.3. History and current orefield exploration The Yueshan intrusion and its surroundings became exploration targets due to the high magnetic anomaly detected by airborne magnetic surveying in 1957, but the first drilling hole (ZK111) in the center of the magnetic anomaly drilled by the Anhui 326 geological team in 1959 did not find any ores because of insufficient depth. In 1960, the third drillhole (ZK113) found high grade Fe–Cu skarn at the depths of 382.06 to 443.73 m, demonstrating the potential mineral resource in this orefield. Then, during 1960 to 1978, 1981 to 1988 and 1984 to 1992, three major exploration programs were carried out. The later project targeted the deeper orebodies and increased less ore reserves than the earlier (Table 1). The history and current situation of exploration in this orefield indicated that further exploration strategies have to deal with the following difficulties. (1) The previous ore deposit models can explain why the ores are located along the contact zones of the

intrusion, rather than why the ores are only located in some specific places along the contact zone. This means that the previous ore deposit models are ineffective in selecting drilling targets. (2) The magnetic anomalies were previously successfully used for the selection of drilling targets, but they have also become less and less effective, because the undiscovered potential orebodies at depth can make the magnetic anomalies too weak to be detected on the surface and to be distinguished from the anomalies induced by the shallow orebodies. (3) The previous geophysical data did not reveal the deep architectures of the orefield, so that they cannot provide detailed enough information for selecting drilling targets for deep ore exploration. In order to facilitate the predictive discovery of deep orebodies, we must understand the deep architecture of the orefield in great detail by making use of advanced geophysical technologies. This will allow more accurate interpretations about the geodynamic constraints on the orebody localization.

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

253

Fig. 3. Geological map of the Anqing orefield (a) and cross-section of profile A–B in the Anqing mine (b), showing the deep architecture deduced from the inversion of magnetotelluric data. Data compiled from Anhui 326 Geological Team (1978, 1988, 1992).

3. Geophysical information of the deep architecture

Fig. 4. 3D shape of the Yuenshan intrusion, modeled from the direct current resistivity sounding data.

The deep architecture is a critical precondition for understanding the mineralization system and predicting potential orebodies at depth. In the Anqing orefield, the most significant factor of the deep architecture is the deep contact zone of the diorite intrusion with the sedimentary carbonate. In order to gain more detailed information about the deep architecture, we conducted a magnetotelluric survey by making use of the Stratagem EH4 method developed by Geometrics (Geometrics, 2000). The Stratagem EH4 is an advanced electromagnetic method. It is extensively used in the mine exploration (e.g., Shen et al., 2008). In the southern part of the Anqing mine, we have carried out a magnetotelluric survey on a grid spacing of 200 × 40 m, mainly for detecting the deeper-extension of the south contact zone that had been revealed by drilling holes. The orientation of the profile is N–S (Fig. 6). The apparent resistivities of different frequencies are

254

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

Fig. 5. 3D relationship of the major orebodies with the eastern branch of the Yueshan intrusion. (a) Entire view of contact zone and orebody; (b) cross-section along L1–L2 and view toward east, showing typical “step-shaped” contact zone and orebody; (c) cross-section along M1–M2 and view toward east, showing typical “tongue-shaped” contact zone and orebody.

Fig. 6. Local geological sketch map of the Anqing mine, showing locations of the magnetotelluric profiles and drilling profile 0–0 (Fig. 9).

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

255

Table 1 Exploration stage and ore reserve increase in the Anqing deposit. Year

1960 to 1978 1981 to 1988 1984 to 1992

Number of drillholes

172 28 33

Total drillhole length (m)

76,725.1 22,420.1 28,281.3

Aver. length per drillhole (m) 446.1 800.7 857

Reserve increase (104t)

Reserve increase per meter of drillhole (t/m)

Copper

Iron ore

Copper

Iron ore

42.39 2.025 7.558

3060 891.08 626.21

5.5 0.9 2.67

398.9 397.45 221.42

Target (depth/m)

W MaAnShan (− 200 to − 500) MaTouShan (− 300 to − 800) E MaAnShan (− 500 to − 900)

Data from Anhui 326 Geological Team (1978, 1988, 1992).

inversed into resistivity section using the IMAGEM-2D analysis software that was also developed by Geometrics. According to the subsurface resistivity and its spatial variations, we can basically identify the major petrologic units and major structures. The highest resistivity, usually N10,000 Ω·m, is definitely a response from the carbonate rocks, while the moderate high resistivity, usually between 2000 and 10,000 Ω·m, reflects the diorite rocks. The contact zone is usually a high gradient transition zone between the highest resistivity domain and the moderate high resistivity domain. All rocks, if located near the surface and fractured, must have a lower resistivity or even the lowest resistivity, because of the influence of water. The resistivity sections reflect the southern contact zone and support the conclusion that the deep-extension of the contact zone repeats the shape of the upper part. The “deep step” occurs under the “upper step”, while the “deep tongue” occurs under the “upper tongue” (Figs. 7 and 8). This reasoning provides very useful information for constructing a geodynamic model of the mineral system, so that it can be used to predict the presence of deep orebodies.

Fig. 7. Resistivity section under profile A40, showing geological interpretation. Codes for all lithostratigraphic units are explained in Fig. 2.

4. Geological information on the geodynamic system of oreformation The wall rocks of the Yueshan intrusion were folded, but the shape and location of the orebodies are not associated with the fold. The Fe– Cu mineralization can occur in brittle fractures within the folded wall rocks (Figs. 3 and 9). This indicates that mineralization is not controlled by the fold and must have taken place after the folding event. Ore veins occur in the fractures and faults within the diorite (Fig. 9), but barren diorite clasts occur in a matrix composed of sulfide-bearing magnetite ores. This fact indicates that the Cu- and Fe-sulfides must have been deposited when the magmatic intrusion had solidified. The thermal energy released by the intrusion during its cooling probably drove hydrothermal circulations (Zhao et al., 2003). The igneous intrusions are controlled by the thrust-fold belt formed during the Indosinian (Triassic) orogeny, while the cooling ages of the intrusions can range from 138 to 142 Ma (Early Cretaceous), when the tectonic regime changed from compression to extension (Dong and Qui, 1993; Mao et al., 2006; Zhang et al., 2008). This implies that the hydrothermal system was formed during cooling of the intrusions in a tensional tectonic setting. Copper orebodies are definitely associated with skarns, but their relationship is very complex. Major skarn minerals such as garnet and diopside must have been formed before the formation of Cu-sulfides and Fe-oxides, because the sulfides and oxides usually form veinlets in the skarn that usually occurs as clasts in the ores (Fig. 10). Statistically, only a small proportion of skarns coexists with Cu-sulfide or Fe-oxide orebodies. This indicates that the Cu and Fe mineralization is not the inevitable results of the skarnification process (Chang et al., 1991; Liu et

Fig. 8. Resistivity section under profile A46, showing geological interpretation. Legends are the same as Fig. 7. Codes for all lithostratigraphic units are explained in Fig. 2.

256

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

the formation of the orebodies. This is because the variation of a contact zone shape can directly cause the variation of local physical conditions. Most orebodies have sharp boundaries, while the adjacent carbonate wall rocks have no alteration (Fig. 11). This means that, although the formation of the skarn is exclusively due to chemical reaction processes, the formation of the Cu or Fe–Cu orebodies must involve a strong physical process to form the necessary dilation zone for ore deposition. The unique irregular boundary of the orebody (Fig. 11) indicates that the dilation zone for hosting ores was formed by a tensional process. The temperature of the inclusions in hydrothermal ore systems is 220 to 585 °C (Zhou et al., 2007), suggesting that during ore deposition, there was a significant drop in temperature. Stable isotope characteristics (H, O, S, Si, He and Ne) indicate that the ore components might mainly come from the intrusion, with some additional materials from the sedimentary rocks (Zhou et al., 2004, 2007). This implies that a rapid temperature drop and the fluid mixing may have played an important role in the ore deposition processes.

5. Computational modeling of coupled physical processes related to ore-formation 5.1. Mathematical relation and computer code

al., 2008), although the skarn provides a favorable condition for the Cu and Fe ore formation. The close relationship of the orebodies to the shape of the contact zone suggests that physical conditions such as stress, strain and temperature might have played an important role in

Computational geodynamic modeling uses numerical methods to solve the dynamic equations of a geodynamic system. It requires robust computer codes that can reproduce the full range of real rock behaviors and incorporate any strong feedback coupling between different processes. In this investigation, we used FLAC (Fast Lagrangian Analysis of Continua) to simulate deformation, heat transfer and fluid flow during the syn-deformation cooling of ore-related intrusion. FLAC is a two-dimensional explicit finite-difference code for simulating the mechano-thermo-hydrological behavior of a continuous porous medium (Itasca Consulting Group, 2002). Compared with finite-element models, explicit finite-difference models have the following obvious advantage: there is no ‘global stiffness’ matrix that needs to be inverted during each time-step, although the time-step must be small enough to ensure that the physical wave velocity never exceeds the “calculated wave velocity”. FLAC updates coordinate positions using the displacements computed from the previous time-step, which can be easily done as there is no global stiffness matrix to invert. The grid, therefore, is displaced along with the material it represents (Cundall and Board, 1988; Strayer et al., 2001).

Fig. 10. Photograph in reflected light showing chalcopyrite and pyrite within veinlets in a cracked garnet skarn. Brecciated garnet skarn occurs as clasts, indicating that the garnet formed before the sulfide ores, Ga = garnet, Py = pyrite, Cp = chalcopyrite.

Fig. 11. The irregular boundary of orebody I and ore-bearing fractured marble in the Anqing mine.

Fig. 9. Drill section of profile 0–0 (located in Fig. 6), showing that the intrusion is control by the fold and the mineralization is associated with fracturing in the intrusion. The codes for all lithostratigraphic units are explained in Fig. 2.

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

In our FLAC models, the intrusion and its wall rocks are treated as Mohr–Coulomb materials. The coupled geodynamics are governed by the following equations: f

qi ¼ −

 kaij ∂  P−ρf gj xj μ ∂xj

ð1Þ

  T f s ∂T qi ¼ − ϕλij þ ð1−ϕÞλij ∂xj

ð2Þ

∂q f ∂ζ f ¼ − i þ qv ∂xi ∂t

ð3Þ

  ∂qT f s ∂T T f ∂T ϕρf Cv þ ð1−ϕÞρs Cv − i ¼ qv −q i ∂t ∂xi ∂xi

ð4Þ

ρ

d˙u i ∂σij ¼ þ ρgi dt ∂xj

∂εTij ∂t

¼ αT

∂T δ ∂t ij

  ∂P ∂ζH ∂ε ∂T −α v þ β ¼M ∂t ∂t ∂t ∂t

ð5Þ

257

5.3. Model specification The essential difference between these two models is their architecture, not their composition. Therefore, same components in these two models have same properties. The related hydrological, mechanical and thermal properties are shown in Table 2. They were determined mainly on the basis of petrological compositions using test data for the physical properties of rocks (Schön, 1998) and some data in the “FLAC (4.0 version) User's Guide” (Itasca Consulting Group, 2002). Some properties show a very wide range of variation. The data selected for use in the models were based mainly on the consideration of: (1) the variation trend of the property with the mineral assemblage, crystallinity, fabric tightness, size and shape of grains, and (2) the comparison of the modeling results obtained using different data properties with the geological observation. Some properties, such as permeability and porosity, must be changed during deformation. We simply use volumetric strain as an indicator for increasing or decreasing porosity. Their relationship is expressed as follows (Itasca Consulting Group, 2002; Zhao et al., 1999):

ð6Þ

ϕ ¼ 1−ð1−ϕ0 Þ=ð1 þ εv Þ:

ð7Þ

During computational modeling, permeability was changed as a function of porosity using the Carman–Kozeny equation (Nield and Bejan, 1992; Zhao et al., 1999):

where qif and qiT are the fluid specific-discharge vector and the heatflux vector respectively; kija is the apparent mobility coefficient, which is a function of permeability (kij) and saturation (s) as kija = k2 ijs (3 − 2s); μ is the dynamic viscosity of the pore fluid; P is the pressure of the pore fluid;λijf and λijs are the thermal conductivity tensors of the pore fluid and solid, respectively; T is temperature; ρs and ρf are the densities of the solid and pore fluid, respectively; gj is the component of gravitational acceleration in the xj direction; ζ is the variation of pore fluid volume per unit volume of the porous material; qvf and qvT are the volumetric fluid source and thermal source, respectively; Cvf and Cvs are the specific heats of the pore fluid and solid, respectively; σij is the stress tensor of the solid; ρ = (1 − φ)ρs + φρf is the bulk density of the porous medium; φ is the porosity; ˙u i is the velocity component in the xi direction; εijT is the thermal strain tensor; ζH is the variation of fluid content; εv is the volumetric strain; M is the Biot modulus; δij is the Kronecker delta; α is the Biot coefficient, and β is the volumetric thermal expansion coefficient. Eqs. (1) and (2) are the Darcy law describing pore fluid flow and the Fourier law describing heat transfer, respectively; Eqs. (3), (4) and (5) describe the conservation of mass, energy and momentum, respectively; and Eqs. (6) and (7) describe the coupled MTH constitutive relations. 5.2. Construction of models Generally, the synthesis of the surface plane and typical cross section of the Anqing Fe–Cu deposit (Fig. 3), the 3D model of the Yueshan intrusion and related orebodies (Figs. 3 and 4), and the observed geological and geophysical constraints to the deep architecture (Figs. 7 and 8) are very useful information for constructing a conceptual model of mineralization in this orefield. Together with consideration of the geodynamic evolution associated with the Anqing deposit and the regional geology, the above information is used to construct some two 2D models for simulating the coupled MTH geodynamic processes during the syntectonic cooling of the Yueshan intrusion. Model A (Fig. 12a) is used to simulate the “tongue-shaped” contact zone of the intrusion, while Model B (Fig. 12b) is used to simulate the “step-shaped” contact zone of the intrusion. Both two models represent a 9000 m long and 4000 m deep cross-section. Details of the geometrical architecture of these two models are given in Fig. 12.

k¼

k0 ð1−ϕ0 Þ2 ϕ3 ϕ30 ð1−ϕÞ2

ð8Þ

ð9Þ

where φ0 and k0 are the initial porosity and initial permeability, respectively. 5.4. Initial and boundary conditions In these two models, the initial temperature of the top surface is set at 25 °C and kept constant; a temperature gradient of 20 °C/km is applied to the sedimentary section; and the initial temperature of the intrusion is set at 600 °C. All pore spaces are initially saturated by fluids. The initial pore fluid pressure of the sedimentary rocks and skarn is set to be hydrostatic pressure, while the initial pore fluid pressure of the intrusion is set to be lithostatic pressure, based on the general model of Fournier (1999). These two models are also subjected to horizontally extensional stretching deformation with the boundary velocity of 1.73 × 10 − 10 ms − 1 symmetrically applied at both sides. The boundaries are insulated for heat and impermeable for pore fluids. These conditions were chosen by the following considerations. (1) Due to the limitation of the computer code, our numerical experiments cannot simulate the phase change process from liquid magma to crystalline rocks, but can simulate the syntectonic cooling process of the intrusion after solidification, which is closely associated with ore deposition. The initial temperature of the intrusion in the experiments must thus be a little lower than the crystallizing temperature of the intrusion and a little higher than the highest temperature of ore-forming fluids. The crystallizing temperature of the intrusion is 780 °C (Zhou and Yue, 1995), and the highest temperature of the ore-forming fluids estimated from the fluid inclusion is 585 °C (Zhou et al., 2007). Thus, 600 °C is an acceptable initial temperature for the intrusion in the experiments. (2) Limited by the FLAC code, our numerical experiments only simulate single phase fluid flow in the porous medium. The major composition of the fluids is liquid water. Thus, the initial saturation is set as 1.

258

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

Fig. 12. Model A (a) and Model B (b), showing their geometry, constitutive components and boundary conditions. Both models are subjected to horizontally extensional stretching deformation with the boundary velocity of 1.73 × 10− 10 ms− 1 symmetrically applied at both sides. Codes for all lithostratigraphic units are explained in Fig. 2.

(3) The boundary velocity is a little faster than the stretching velocity of the regional crust during the early Cretaceous. 5.5. Modeling results and implications for ore formation geodynamics The computational modeling experiments of Models A and B produced the same results that can provide some significant indications for ore formation geodynamics. The dilatant deformation, pore fluid

flow pattern and temperature spatial variation obtained from the experiments show remarkable regular patterns and close association with the location of the orebodies and geological architectures. Dilation zones are the most evident deformation in the modeling experiments (Figs. 13 and 14). Except for the dilation zones near the top and side boundaries, where the results may have no metallogenic implication because of the artificial boundary effects (Zhao, 2009), the significant dilation zones are distributed unevenly along the intrusive contact.

Table 2 Parameters of the Anqing model. Component (rock unit)

Density (kg/m3)

Bulk modulus (1010 Pa)

Shear modulus (1010 Pa)

Tensile strength (106 Pa)

Cohesion (106 Pa)

Friction angle

Dilation angle

Permeability (10− 12 m2)

Thermal conductivity (W·m− 1·K− 1)

J T3l T2t T2y2 T2y1 T1n T1y-T1h P2 C-P1 Diorite Skarn

2500 2520 2560 2650 2680 2620 2630 2615 2625 2600 2590

1.9 2.5 3.2 2.0 1.8 4.8 2.8 2.2 5.8 2.95 1.9

0.72 0.82 1.92 0.92 1.4 2.6 2.1 1.2 3.6 1.97 0.92

1.3 1.8 2.2 1.0 2.88 3.2 2.28 1.8 3.2 2.2 0.32

2.6 3.8 4.0 2.8 6.2 5.8 4.2 3.2 4.8 3.8 0.98

15 25 32 30 35 38 28 10 35 25 35

4 3 4 5 25 18 21 5 19 8 30

30 30 28 25 89 4.2 19 8.6 3.8 8.5 55

1.0 1.8 2.2 2.3 5.3 3.2 2.3 4.5 3.0 2.0 2.6

Codes for rock unit are as same as in Fig. 2.

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

In Model A, the dilation zones are located in the “tongue-tips” of the “tongue-shaped” contact zone, while in Model B, the dilation zones are located in the steep segments of the “step-shaped” contact zone. Both models show that the significant dilation zones are located in the south contact zone rather than the north contact zone of the intrusion. The close association of the dilation deformation with the abrupt variation of the properties and the contact shape of the intrusion indicates that the rheological properties, the thermal contrasts and the contact shape can play a controlling role in the coupled MTH process. The dilation zones produced by the computational modeling are spatially associated with the orebodies in the orefield. By analyzing their association and the modeling conditions, some understandings of the geodynamic constraints on the ore localization can be gained as follows. (1) All existing major orebodies are strictly located in the dilation zones along the south contact boundary of the intrusion, and no orebody has been discovered in the contact zone where no dilatant deformation was produced (Figs. 13 and 14). This means that the uneven distribution of orebodies along the contact zone of the intrusion can be attributed to the uneven distribution of dilatant deformation. (2) Some dilation zones near the top and side boundaries are definitely barren, but orebodies have been discovered in drilling holes within all dilation zones along the contact boundary of the intrusion. This demonstrates that the contact zone has played an important role in the ore formation of this orefield.

259

(3) Because no chemical processes were involved in the computational modeling and no faults were constructed in the model, the dilation zones related to the orebodies are absolutely the results of the coupled MTH processes. This indicates that the MTH processes can play an important role in the localization of the skarn ores, although the skarn ores might be produced as a result of chemical-thermal processes. (4) When the models are subjected to zero boundary velocity or shortening deformation with the boundary velocity of 1.73 × 10 − 10 ms − 1 symmetrically applied at both sides, the models cannot produce results as mentioned above. In this case, no dilation zones occur even though there exist orebodies. This implies that the extensional setting is important for forming the favorable mechanical conditions of ore formation. (5) The dilation zones along the contact boundary of the intrusion are not only the focusing center of the pore fluids from the sedimentary rocks and intrusion (Figs. 13 and 14), but also the sharpest drop region of temperature and the region of relative low pore fluid pressure (Figs. 15 and 16). These conditions are all favorable for ore deposition. The focusing of pore fluids from different sources can result in the mixing reaction, an effective mechanism for ore deposition. The sharp temperature drop of magmatic hydrothermal fluids can cause the sharp decreasing of metal solubility, and result in the deposition of metal minerals. The decrease of pore fluid pressure can cause the effervescence and phase separation of high-pressured magmatic hydrothermal fluids, resulting in the deposition of

Fig. 13. Deformation, temperature and pore fluid flow results of Model A, showing Darcy velocities (arrows), isotherms and total volumetric strain contours at 9200 years, the maximum pore fluid flow velocity is 3.262 × 10− 6 m/s: (a) entire view; and (b) zoomed-in view.

260

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

Fig. 14. Deformation, temperature and pore fluid flow results of Model B, showing Darcy velocities (arrows), isotherms and total volumetric strain contours at 6000 years, the maximum pore fluid flow velocity is 2.042 × 10− 6 m/s: (a) entire view; and (b) zoomed-in view.

metal minerals. It is these favorable conditions that result in the localization of orebodies in the dilation zones along the contact boundary of the intrusion. 6. Contributions of geodynamic modeling to targeting deeper orebodies The modeling results show that there are major dilation zones at depth in the Anqing deposit (Figs. 13 and 14). This means that these zones are favorable locations for finding orebodies. The appropriate targets for exploring the potential orebodies can be arranged at the following two locations: (1) the “tongue-tips” of the deeper “tongue-shaped” contact zone in the west part of the Anqing mine, and (2) the “steep segments” of the deeper “step-shaped” contact zones in the east part of the Anqing mine. The modeling results also demonstrate that no orebody exists in the north contact zone of the intrusion, because no dilation zone was produced there. Thus, the deep exploration should target the south contact zone rather than the north contact zone of the intrusion. The controls of dilation zones on the location of orebodies along the contact boundary of the intrusion suggest that exploration at

deeper levels should target the dilation zones close to the deep contact boundaries of the intrusion. Because our models are constructed on the basis of 2D conceptual sections, every dilation zone produced in the modeling process cannot represent an exact location in the field. Thus, the modeling results are used to generate exploration target concepts rather than exact drilling locations. The targets should be the deeper “tongue-tips” and the deeper “steep segments” within the contact zones of the intrusion. The exact locations of the deeper “tongue-tips” and the deeper “steep segments” within the contact zones of the intrusion depend on the geophysical data. Therefore, the reliability and accuracy of geophysical data are also critical for determining drilling locations. In fact, based on the modeling results, a new orebody was discovered from two drilling holes in profile A46 (Fig. 17) during an ongoing deep drilling program in the depth of the Anqing deposit. One drill hole located at the northern contact zone of the intrusion did not find orebody, and three other holes also did not find orebody at the southern contact zone of the intrusion, as a consequence of the inaccurate prediction of the location of the deeper tongue-tips and deeper segments of the contact zones. All these results demonstrated that: (1) the computational modeling of coupled geodynamic processes

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

261

Fig. 15. Variation curves of the volumetric strain, pore fluid pressure and temperature along line A1–A2 in Model A (located on Fig. 13a). The gray bar indicates the location of pore fluid focusing dilation zone along the contact boundary of the intrusion.

can play an important role in facilitating the discovery of deep-sited blind orebodies in a well-explored district such as the Anqing orefield, and (2) the details and accuracy of the architecture obtained from geophysical and geological data are also critical for a successful drilling program.

Fig. 17. Cross-section of profile A46 in the Anqing mine, showing the orebody discovered in the predicted target (deep dilation zone).

7. Conclusions The Fe–Cu Skarn deposits in the Anqing orefield were formed during cooling of the Yueshan diorite intrusion when the crust was

subjected to extensional deformation. The geological characteristics of the Anqing deposit indicate that the ores were deposited in tensional structural spaces along the intrusive contact zone, and the location of ores is closely related to the shape of the contact zone. The computational modeling experiments based on geological and geophysical data demonstrate that the coupled geodynamic processes can play a controlling role in the location of orebodies. Therefore, the dilation zones close to the contact boundary of the intrusion can control the location of orebodies, which is a direct result of the coupled MTH processes during syntectonic cooling of the intrusion. The skarns do not control the localization of the orebodies. This recognition provides useful information for selecting exploration targets. Specifically, deep exploration should target the deep dilation zones close to the contact zones of the intrusion. Based on the computational modeling prediction, an orebody has been discovered in the deep dilation zone of the Anqing orefield. This demonstrates that the computational modeling is a promising tool for understanding the metallogenic geodynamics and for facilitating the deep exploration of hidden orebodies that are related to intrusions. Acknowledgments

Fig. 16. Variation curves of the volumetric strain, pore fluid pressure and temperature along line B1–B2 in Model B (located on Fig. 14a). The gray bar indicates the location of pore fluid focusing dilation zone along the contact boundary of the intrusion.

The work presented here is financially supported by a grant (No. 40772195) from the Natural Science Foundation of China (NSFC). We acknowledge the Tongling Nonferrous Metal Group Inc. for support of the field investigation and geophysical survey. We also would like to thank Dr. Z.Z. Xi, Mr. X.H. Yuan, Mr. Z.M. Shu and Mr. C. Zhou for their assistance in the geophysical and geological investigation. Special thanks are given to Prof. Nigel Cook, Dr. Franco Pirajno, Dr. González-Álvarez and an anonymous reviewer for their valuable comments on our early version of this paper.

262

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263

References Anhui 326 Geological Team, 1978. Report of General Exploration in Western Ma-AnShan, Anqing Copper Deposit. Anhui Provincial Geological Bureau, Hefei, Anhui. 215 pp. (in Chinese). Anhui 326 Geological Team, 1988. Report of General Exploration in Ma-Tou-Shan, Anqing Copper Deposit. Anhui Provincial Geological Bureau, Hefei, Anhui. 203 pp. (in Chinese). Anhui 326 Geological Team, 1992. Report of General Exploration in Eastern Ma-AnShan, Anqing Copper Deposit. Anhui Provincial Geological Bureau, Hefei, Anhui. 278 pp. (in Chinese). Anhui Provincial Geological Bureau, 1987. The Regional Geology of Anhui Province. Geological Publishing House, Beijing. 721 pp. (in Chinese, with English abstract). Candela, P.A., Philip, M., Piccoli, P.M., 2005. Magmatic processes in the development of porphyry-type ore systems. Economic Geology 100th Anniversary volume, pp. 25–38. Carranza, E.J.M., Sadeghi, M., 2010. Predictive mapping of prospectivity and quantitative estimation of undiscovered VMS deposits in Skellefte district (Sweden). Ore Geol. Rev. 38, 219–241. Chang, Y.F., Liu, X.P., Wu, Y.C., 1991. The Copper–Iron Belt of the Low and Middle Reaches of the Yangtze River. Geological Publish House, Beijing. 359 pp. (in Chinese with English abstract). Chi, G., Savard, M.M., 1998. Basinal fluid flow models related to Zn–Pb mineralization in the southern margin of the Maritimes Basins, Eastern Canada. Econ. Geol. 93, 896–910. Cundall, P.A., Board, M., 1988. A microcomputer program for modelling large-strain plasticity problems. In: Swoboda, C. (Ed.), Numerical Methods in Geomechanics: Proceedings, 6th International Conference on Numerical Methods in Geomechanics. Balkema, Rotterdam, pp. 2101–2108. Dong, S.W., Qiu, R.L., 1993. Tectonism and Magmatism in the Anqing-Yueshan Area. Geological Publishing House, Beijing. 154 pp. (in Chinese with English abstract). Etheridge, M.A., Henley, R.W., 1997. Making models matter. In: Price, G.P. (Ed.), Geodynamics and Ore Deposits Conference Abstracts, Australian Geodynamics Cooperative Research Centre, Nedlands, W.A., Australia, pp. 4–5. Ford, A., McCuaig, T.C., 2010. The effect of map scale on geological complexity for computer-aided exploration targeting. Ore Geol. Rev. 38, 156–167. Fournier, R.O., 1999. Hydrothermal processes related to movement of fluid from plastic into brittle rock in the magmatic-epithermal environment. Econ. Geol. 94, 1193–1211. Franklin, J.M., Gibson, H.L., Jonasson, I.R., Galley, A.G., 2005. Volcanogenic massive sulfide deposits. Economic Geology, 100th Anniversary Volume, pp. 523–560. Garven, G., Freeze, R.A., 1984. Theoretical analysis of the role of groundwater flow in the genesis of stratabound ore deposits: mathematical and numerical model. Am. J. Sci. 284, 1085–1124. Geometrics, 2000. Operation Manual for Stratagem Systems Running IMAGEM, Version 2.16. http://www.geometrics.com/Stratagem/stratagem.html. Hobbs, B.E., Zhang, Y., Ord, A., Zhao, C., 2000. Application of coupled deformation, fluid flow, thermal and chemical modeling to predictive mineral exploration. J. Geochem. Explor. 69–70, 505–509. Hronsky, J.M.A., Groves, D.I., 2008. Science of targeting: definition, strategies, targeting and performance measurement. Aust. J. Earth Sci. 55, 3–12. Itasca Consulting Group, 2002. FLAC User's Guide. Itasca Consulting Group, Inc, Minneapolis. Kerrich, R., Richard, J., Goldfarb, R.J., Jeremy, P., Richards, J.P., 2005. Metallogenic provinces in an evolving geodynamic framework. Economic Geology 100th Anniversary Volume, pp. 1097–1113. Knox-Robinson, C.M., 2000. Vectorial fuzzy logic: a novel technique for enhanced mineral prospectivity mapping, with reference to the orogenic gold mineralisation potential of the Kalgoorlie Terrane, Western Australia. Aust. J. Earth Sci. 47, 929–941. Knox-Robinson, C.M., Wyborn, A.I., 1997. Towards a holistic exploration strategy: using Geographic Information Systems as a tool to enhance exploration. Aust. J. Earth Sci. 44, 453–463. Kreuzer, O.P., Etheridge, M.A., Guj, P., McMahon, M.E., Holden, D.J., 2008. Linking mineral deposit models to quantitative risk analysis and decision-making in exploration. Econ. Geol. 103, 829–850. Liu, L.M., 2007. Predictive capability of metallogenic theory and its facilitating approaches. Earth Sci. Front. 14 (5), 82–91 (in Chinese with English abstract). Liu, L.M., Peng, L.M., 2003. Prediction of hidden ore bodies by synthesis of geological, geophysical and geochemical information based on dynamic model in Fenghuangshan ore field, Tongling district, China. J. Geochem. Explor. 81, 81–98. Liu, L.M., Peng, S.L., 2005. Key strategy for predictive exploration in mature environment: model innovation, exploration technology optimization and information integration. J. Cent. South Univ. Technol. 12, 186–191. Liu, L.M., Zhang, Y., 2007. Numerical modeling of the coupled mechanical and hydrological processes during deformation and mineralization in the Mount Isa block, Australia. Resour. Geol. 57, 283–300. Liu, L.M., Yang, G.Y., Peng, S.L., 2005a. Numerical modelling of coupled geodynamical processes and its role in facilitating predictive ore discovery: an example from Tongling, China. Resour. Geol. 55, 21–31. Liu, L.M., Deng, J.N., Peng, S.L., Shu, Z.M., Shao, Y.J., Yang, H.Y., 2005b. Risk of mineral exploration and predictive exploration strategy for reducing the risk. Geol. Bull. China 24, 1016–1021 (in Chinese with English abstract). Liu, L.M., Shu, Z.M., Zhao, C.B., Wan, C.L., Cai, A.L., Zhai, Y.L., 2008. The controlling mechanism of ore formation due to flow-focusing dilation spaces in skarn ore

deposits and its significances for deep-ore exploration: Examples from the Tongling-Anqing district. Acta Petrol. Sin. 24, 1848–1856 (in Chinese with English abstract). Mair, J.L., Ojala, V.J., Salier, B.P., Groves, D.I., Browns, M., 2000. Application of stress mapping in cross-section to understanding ore geometry, predicting ore zones and development of drilling strategies. Aust. J. Earth Sci. 47, 895–912. Mao, J.W., Wang, Y.T., Lehmann, B., Yu, J.J., Du, A.D., Mei, Y.X., Li, Y.F., Zang, W.S., Stein, H.J., Zhou, T.F., 2006. Molybdenite Re–Os and albite 40Ar/39Ar dating of Cu–Au–Mo and magnetite porphyry systems in the Yangtze River valley and metallogenic implications. Ore Geol. Rev. 29, 307–324. Meinert, L.D., Dipple, G.M., Nicolescu, S., 2005. World skarn deposits. Economic Geology 100th Anniversary Volume, pp. 299–336. Nield, D.A., Bejan, A., 1992. Convection in Porous Media. Spring-Verlag, New York. 408 pp. Oliver, N.H.S., Ord, A., Valenta, R., Upton, P., 2001. Deformation, fluid flow, and ore genesis in heterogeneous rocks, with examples and numerical models from the Mount Isa district, Australia. Rev. Econ. Geol. 14, 51–71. Ord, A., Hobbs, B.E., Zhang, Y., Broadbent, G.C., Brown, M., Willetts, G., Sorjonen-Ward, P., Walshe, J.L., Zhao, C., 2002. Geodynamic modeling of the Century deposit, Mt Isa Province, Queensland. Aust. J. Earth Sci. 49, 1011–1039. Pan, Y., Dong, P., 1999. The lower Changjiang (Yangze/Yangtze River) metallogenic belt, Easter Central China: intrusion- and wall rock-hosted Cu–Fe–Au, Mo, Zn, Pb, Ag deposits. Ore Geol. Rev. 15, 177–242. Partington, G., 2010. Developing models using GIS to assess geological and economic risk: an example from VMS copper gold mineral exploration in Oman. Ore Geol. Rev. 38, 197–207. Porwal, A., Carranza, E.J., Hale, M., 2003. Knowledge-driven and data-driven fuzzy models for predictive mineral potential mapping. Nat. Resour. Res. 12, 1–25. President's Information Technology Advisory Committee, 2005. Computational Science: Ensuring America's Competitiveness. http://www.nitrd.gov/pitac/reports/ 20050609_comoutational/computational.pdf. Price, G.P., Stoker, P., 2002. Australian Geodynamics Cooperative Research Center's integrated research program delivers a new minerals exploration strategy for industry. Aust. J. Earth Sci. 49, 595–600. Raffensperger, J.P., Garven, G., 1995. The formation of unconformity-type uranium ore deposits: coupled hydrochemical modeling. Am. J. Sci. 295, 639–696. Schaubs, P.M., Zhao, C., 2002. Numerical models of gold-deposit formation in the Bendigo–Ballarat zone, Victoria. Aust. J. Earth Sci. 49, 1077–1098. Schön, J.H., 1998. Physical Properties of Rocks: Fundamentals and Principles of Petrophysics. Pergamon-Elsevier, Oxford. 583 pp. Shen, P., Shen, Y.C., Liu, T.B., Li, G.M., Zeng, Q.D., 2008. Prediction of hidden Au and Cu– Ni ores from depleted mines in Northwestern China: four case studies of integrated geological and geophysical investigations. Miner. Deposita 43, 499–517. Sorjonen-Ward, P., Zhang, Y., Zhao, C., 2002. Numerical modeling of orogenic processes and gold mineralisation in the southeastern part of the Yilgarn Craton, West Australia. Aust. J. Earth Sci. 49, 935–964. Strayer, L.M., Hudleston, P.J., Lorig, L.J., 2001. A numerical model of deformation and fluid-flow in an evolving thrust wedge. Tectonophysics 335, 121–145. Wang, X.C., Zhou, Y.C., 1995. Geological characteristics and origin of Anqing Cu–Fe deposit. Anhui. Geol. Prospect. 31 (3), 16–23 (in Chinese with English abstract). Wyborn, L.A.I., Gallagher, R., Jagodzinski, E.A., 1994. A conceptual approach to metallogenic modelling using GIS: examples from the Pine Creek Inlier. Proceedings of a Symposium on Australian Research in Ore Genesis, Glenside, South Australia, Australian Mineral Foundation, pp. 15.1–15.5. Yang, J., Large, R.R., Bull, S., Scott, D.L., 2006. Basin-scale numerical modeling to test the role of buoyancy-driven fluid flow and heat transfer in the formation of stratiform Zn–Pb–Ag deposits in the northern Mount Isa Basin. Econ. Geol. 101, 1275–1292. Zhai, Y.S., Xiong, Y.L., Yao, S.Z., Lin, X.D., 1996. Metallogeny of copper and iron deposits in the Eastern Yangtze Craton, east-central China. Ore Geol. Rev. 11, 229–248. Zhang, Y., Lin, G., Roberts, P., Ord, A., 2007. Numerical modelling of deformation and fluid flow in the Shuikoushan district, Hunan Province, South China. Ore Geol. Rev. 31, 261–278. Zhang, L.J., Zhou, T.F., Fan, Y., Yuan, F., 2008. SHRIMP U–Pb zircon dating of Yueshan intrusion in the Yueshan ore field, Anhui, and its significance. Acta Petrol. Sin. 24, 1725–1732 (in Chinese with English abstract). Zhao, C., 2009. Dynamic and Transient Infinite Elements: Theory and Geophysical, Geotechnical and Geoenvironmental Applications. Springer, Berlin. 259 pp. Zhao, C., Hobbs, B.E., Mühlhaus, H.B., 1999. Effects of medium thermoelasticity on high Rayleigh number steady-state heat transfer and mineralization in deformable fluid-saturated porous media heated from below. Comput. Meth. Appl. Mech. Eng. 173, 41–54. Zhao, C., Lin, G., Hobbs, B.E., Ord, A., Wang, Y., Muhlhaus, H.B., 2003. Effects of hot intrusions on pore-fluid flow and heat transfer in fluid-saturated rocks. Comput. Meth. Appl. Mech. Eng. 192, 2007–2030. Zhao, C., Hobbs, B.E., Hornby, P., Ord, A., Peng, S., 2006. Numerical modelling of fluids mixing, heat transfer and non-equilibrium redox chemical reactions in fluidsaturated porous rocks. Int. J. Numer. Methods Eng. 66, 1061–1078. Zhao, C., Hobbs, B.E., Ord, A., Hornby, P., Peng, S., Liu, L., 2007. Mineral precipitation associated with vertical fault zones: the interaction of solute advection, diffusion and chemical kinetics. Geofluids 7, 3–18. Zhao, C., Hobbs, B.E., Ord, A., Hornby, P., Peng, S., 2008. Morphological evolution of three-dimensional chemical dissolution front in fluid-saturated porous media: a numerical simulation approach. Geofluids 8, 113–127. Zhao, C., Hobbs, B.E., Ord, A., 2009. Fundamentals of Computational Geoscience: Numerical Methods and Algorithms. Springer, Berlin. 241 pp.

L. Liu et al. / Ore Geology Reviews 43 (2011) 249–263 Zhao, C., Hobbs, B.E., Ord, A., 2010. Theoretical and numerical investigation into roles of geofluid flow in ore forming systems: integrated mass conservation and generic model approach. J. Geochem. Explor. 106, 251–260. Zhou, T.F., Yue, S.C., 1995. Studies on the physicochemical conditions of magmatic crystallisation in diorites in the Yueshan area, Anhui Province. J. Hefei Univ. Technol. 18, 175–181 (in Chinese with English abstract).

263

Zhou, T.F., Yuan, F., Yue, S.C., Liu, X.D., 2004. Silicon, helium and neon isotope geochemistry of Cu–Au ore deposit in the Yueshan orefield, Anhui Province. Bull. Mineral., Petrol. Geochem. 20, 385–387 (in Chinese with English abstract). Zhou, T.F., Yuan, F., Yue, S.C., Liu, X.D., Zhang, X., Fan, Y., 2007. Geochemistry and evolution of ore-forming fluids of the Yueshan Cu–Au skarn- and vein-type deposits, Anhui Province, South China. Ore Geol. Rev. 31, 279–303.