Geometry and deformation of the sole of a large obducted ophiolitic unit: insights from surface geology and airborne electromagnetics – Peridotite Nappe, New Caledonia

ABSTRACT

To clarify the serpentinite correlation and kinematics in New Caledonia, we focus on the northwestern boundary of the Massif du Sud and the Mont Do Massif, in the Boulouparis area. On the basis of airborne electromagnetics data, structural analysis, and field data, we propose a significantly different model to the one commonly proposed for Peridotite Nappe emplacement. The interpretation is that serpentinite slivers in basement actually correlate with the Peridotite Nappe instead of older basement terrane serpentinites. Therefore the basal Peridotite Nappe fault is not as planar and simple as previously thought. Some portions of the Peridotite Nappe are involved into major shear zones with the basement by duplexing or internal thrust within the peridotite. Most of the kinematic indicators, in the serpentinite slivers, and in the substrate, show a N-S shortening accommodated with lateral (both dextral and sinistral) senses of shear. Serpentinite plays a major role in this process by accommodating most of the deformation and by partitioning strain between the substrate and the nappe, i.e. between the lower and the upper plate. This model is consistent with the dextral oblique convergence model recently proposed by several authors for the emplacement of the subduction-obduction complex of New Caledonia.

KEYWORDS:

• New Caledonia; ophiolite
• Peridotite Nappe
• serpentinite
• airborne electromagnetics
• obduction

Introduction

The Peridotite Nappe of New Caledonia is an iconic example of the obduction of a large ultrabasic ophiolitic unit over continental crust (Maurizot et al. 2020a). The role of serpentinites in this special geodynamic process of emplacement of dense oceanic lithosphere over light continental crust in convergence zones is crucial. Serpentinites, owing to their rheological properties, particularly a very low frictional strength, are favouring strain localisation and partitioning in major deformation zones (Hirth and Guillot 2013; Guillot et al. 2015). A better knowledge of the distribution, geometry, deformation type, and kinematic of these rocks may help in understanding the emplacement mechanism of such ophiolitic complexes.

In New Caledonia, serpentinites occur in two different structural settings (Figure 1): (i) as a sole at the base of the allochthonous obducted ultrabasic units, or (ii) as many slivers within the (relative) autochthonous substrate (herein generically referred to as the ‘substrate’). There is little consideration and even less interpretation in the literature of the association of these two types of occurrences. The serpentine sole represents, at local scale, a generally shallow dipping zone of strong deformation and variable thickness at the base of the Peridotite Nappe which has been schematically interpreted as a result of a compressional or transcurrent mechanism (Cluzel et al. 2012; Gautier et al. 2016; Quesnel et al. 2016) or of a gravitational and extensional passive process (Lagabrielle and Chauvet 2008). Serpentinite inliers are common features in New Caledonia geology. A geographic information system allows to compute that it is not possible to go 5 km in a straight line on Grande Terre (400 × 40 km) without crossing a serpentinite outcrop. They occur as lobate or elongated, contorted and anastomosed bodies running along main tectonic contacts between or into the different units of the substrate. They are generally interpreted as: (i) slices of deep origin extruded from old ophiolitic units of the basement terranes, either the Mesozoic Boghen Terrane (Maurizot et al. 1985; Maurizot 2001), or Late Paleozoic Koh Ophiolite (Meffre 1995; Meffre et al. 1996), or (ii) slices of the Peridotite Nappe sole pinched down in the substrate during Peridotite Nappe emplacement (Guérangé et al. 1975; Leguéré 1976; Gonord 1977; Paris 1981; Maurizot et al. 1985). The two types can coexist and are not easily distinguishable, given the similarities of facies. Whatever their origin, the presence of slices of mantle-derived material in a crustal environment is not trivial. The presence of these odd geological objects implies considerable tectonic deformation. From the point of view of applied geology, these occurrences have also important inferences. Serpentinite slivers are commonly hosts of naturally occurring asbestos (Baumann et al. 2011), which present an important health risk factor for the surrounding population. From a hydrogeological point of view, they are both drains and barriers that control groundwater flow in fractured aquifers.

Figure 1. Simplified structural map of Grande-Terre (main island of New Caledonia) showing the distribution of the ultrabasic formations. Geological map reference: metadata, can be viewed online at the portal https://georep.nc. This online geographical information system includes the availability of, and progress on, New Caledonia geological maps.

In order to improve our knowledge on the 3D geometry of these serpentinite slivers, an Airborne ElectroMagnetic (AEM) survey was collected in the area of Boulouparis. This area is located on the northwestern edge of the Massif du Sud (the largest unit of the Peridotite Nappe), where numerous serpentinite slivers and their relationships to the basal sole of the nappe are well exposed. AEM surveys have already shown their ability to reliably image the subsurface over large areas and in various and complex environments (Auken, Violette et al., 2009; Vittecoq et al. 2015; Sandersen et al. 2021), through the imaging of the resistivity contrasts; the resistivity has an important dynamic which depend on many parameters (rock composition, degree of weathering and clay content, porosity, fluid saturation and its composition) which makes it useful in various themes. In this paper, the AEM survey is interpreted in the light of field observations and structural analyses. This study presents a new tectonic model in accordance with this new geophysical data imaging the distribution and deformation of serpentinite slivers.

Geological setting

New Caledonia, located in the southwest Pacific (Figure 1), is the emergent northern part of the Norfolk ridge, the northeastern most continental ribbon of Zealandia (Mortimer et al. 2017) that rifted away from eastern Gondwana in the Late Cretaceous. Grande Terre, the ‘main island’ of New Caledonia archipelago, is a mosaic of Paleozoic to Cenozoic terranes. Most of these are exposed in the surveyed area of Boulouparis (Figure 2), which is therefore representative of the geology of the country and should help in assessing AEM usefulness.

Figure 2. Simplified geological map of the surveyed area. Red dashed polygon: AEM area. Geological map reference: metadata, can be viewed online at the portal https://georep.nc. This online geographical information system includes the availability of, and progress on, New Caledonia geological maps.

Three amalgamated, initially arc-related, Late Paleozoic to Mesozoic terranes constitute the Gondwanian basement of Grande Terre (Maurizot et al. 2020b). In the investigated area, are exposed (i) the Koh-Central Terrane, which comprises a series of ophiolitic units (Koh Ophiolite for which only the crustal part is known as mafic cumulates and basaltic lavas) covered by thick and monotonous volcaniclastic rocks (greywackes of the Central Terrane) (Adams et al. 2009) and (ii) the Boghen terranes, a high pressure – low temperature (HP-LT) metamorphic complex (Cluzel and Meffre 2002), which protolith is contemporaneous with the volcanic arc that fed the volcanoclastic deposits of the Central Terrane, and is interpreted as a core complex exhumed amongst the Koh-Central Terrane. In the Boulouparis area, the Koh Ophiolite is represented by the Koua Unit (Meffre 1995), which comprises basalts as pillow lavas and dyke complex. It has been dated as Late Paleozoic in the area of Col de Nassirah by U-Pb on zircon method (Aitchison et al. 1998).

In Grande Terre, the basement is unconformably covered by Late Cretaceous to Eocene sedimentary rocks often referred to as ‘the cover’. The Late Cretaceous syn-rift formations that accompanied the rifting and then drifting of Zealandia are not exposed in the surveyed area. Most of the cover is represented by Paleogene turbiditic and syntectonic sequences deposited in the convergence context that precede obduction. These pre-obduction deposits are interpreted as filling up a foreland basin that propagated towards the southwest, in front of the convergence zone, and are often referred to as the Eocene Flysch. The Eocene subduction-obduction complex of New Caledonia is composed of several units that formerly spread to the NE of the Norfolk Ridge and its margin, and mostly in an oceanic domain, namely the south Loyalty Basin. The complex is composed of four allochthonous terranes or nappes which are presently stacked over the continental Norfolk Ridge, with a general vergence from the NE to the SW. Thrust over the Eocene turbidites, these are: (i) the Montagnes Blanches Nappe (Maurizot 2011) which represent the NE margin of the Norfolk Ridge; (ii) the Poya Terrane (Cluzel et al. 1994) which is the upper part of the oceanic crust of the south Loyalty Basin, scrapped off at the front of the convergence zone; (iii) the Peridotite Nappe (Avias 1967) which is the deep lithospheric mantle of the south Loyalty Basin and (iv) the Eocene Metamorphic Belt which is an equivalent of these three former units, metamorphosed to HP-LT schists and gneiss in the NE dipping subduction zone. With the exception of those last metamorphic rocks, the three first units are well exposed in the Boulouparis area.

Three overthrust ultrabasic units are present in the Boulouparis area: (1) the western end of the Massif du Sud, (2) the tiny Ouitchambo klippe, (3) and the Massif du Mont Do. Massif du Sud and Ouitchambo klippe are part of the Peridotite Nappe. However, the correlation of Mont Do Massif to this same unit is not straightforward. The basal contact of Mont Do is gently dipping over the basement terranes and its Eocene Flysch cover on its SW flank, but displays much more complex relationships with that substrate on its NE flank. Its close association with the Kua Unit of the Koh Ophiolite, raises the question of its connection with either this Paleozoic ophiolite or with the younger Peridotite Nappe. In the first case it would represent the lower ultramafic part of the Late Carboniferous ophiolite, thrust over its crustal sequence; in the second case, it would be a subsidiary klippe of the Peridotite Nappe. Indeed, this is an important issue because the assignment to either Koh or to the Peridotite Nappe remains on the updated multi-scale geological map of New-Caledonia (can be viewed online at the portal https://georep.nc), and needs to be clarified area by area.

Methods

Geology and structure

Mont do

The ultrabasic unit of Massif du Mont Do consist of un – to moderately serpentinised layered harzburgite alternating with dunite. It overlies the basement terranes and the Eocene turbidites at an altitude between ∼300 m and ∼700 m. The basal contact is shallow dipping around much of the massif. The serpentinite sole is 20–100 m thick (Figure 3). Serpentinites of the sole are affected by a pervasive deformation, with centimetre – to metre-scale lenses of undeformed serpentinised peridotites embedded and surrounded by foliated mylonitic serpentinites (Figure 4A,B). At the eastern edge of the Mont Do Massif, serpentinites of the sole are well exposed. A main foliation (S1) N150-160, dipping 20–30° to the NE is observed. S/C’ structures indicate a non-coaxial deformation with a top-to-SW sense of shear (Figure 4A). This main foliation is occasionally cross cut by a second deformation with development of spaced cleavage (S2) (Figure 4B) oriented ∼N150 and dipping 70–90° to the NE.

Figure 3. A, Geological map of the eastern part of the Massif du Mont Do. B, Simplified W-E cross-section.

Figure 4. Deformation of serpentinite sole of the Mont Do massif. A, Serpentinite sole at the eastern edge of the Massif du Mont Do with pervasive foliation. B, Two superimposed deformations in the serpentinite sole: boudins of peridotites parallel to S1 (N56 dipping 42° to the South-East) crosscut by a S2 cleavage (N150 dipping 65° to the north-east); C, metre-size boulders of peridotites enclosed by sheared serpentinites (Aramuru ridge); D, boulders of serpentinized peridotites enclosed by sheared serpentinites (Aramuru ridge).

The Aramuru and Oundamien ridges to the northeastern border of the Massif du Mont Do interestingly show the relationship between serpentinite of the basal sole of the Mont Do Massif and the serpentinite slivers cross cutting the basement. The two ridges consist of massive foliated serpentinites embedding lenses of serpentinised peridotites (Figure 4C,D) separating compartments of weakly deformed metasedimentary rocks (metagreywackes, sandstone) of the Boghen or Koh-Central terranes. Detailed mapping shows that serpentinites from both Oundamien and Aramuru ridges belong to a same and single unit wrapping around a kilometre-scale lens of basement. Most importantly, the junction between the basal sole of the Massif du Mont Do and a steeply dipping narrow sliver of serpentinites that crosscut the basement, are very well exposed, and the connecting area can be clearly observed. Both sole and slivers bear their own foliation, which is roughly parallel to their wall rock. Superimposed deformations are observed in the connecting area. The gently dipping S1 foliation (Figure 4A) associated with the serpentinite of the basal sole is cross cut by the steeply dipping S2 cleavage (Figure 4B) only at the junction of the two entities. Farther down, in the serpentinite sliver, S2 cleavage is transpose to a foliation whose direction is parallel to the wall of that sliver. S2 is absent in the enclosing meta-sedimentary rock of the substrate whose metamorphic foliation is distinct. This occurrence may be interpreted in terms of deformation partitioning. The presence of two different foliation orientations and the superimposition of S2 over S1 only in the connecting area indicate that deformation in sole or slivers may occur rather independently. In other words they are two different tectonic expressions in response to the same compressive deformation.

Between Mont Do and Massif du sud

Col de Nassirah area

To the east of the Mont Do massif, serpentinite slivers crosscut both the basement terranes and the Eocene turbidites, forming anasotomosed lenses of kilometer-scale along strike and of an average of 100 m width (Figure 3). In the basement terranes, basalts are commonly associated with the serpentinites as metre to kilometre-scale lenses into the basement (Figures 3 and 5A). These mafic rocks belong to Koh Ophiolite (Meffre 1995).

As described for Mont Do serpentinites, serpentinite slivers display two deformations. A pervasive foliation (S1), which is deformed by a spaced cleavage (S2) (Figure 5B). S2 is parallel to the wall of the serpentinite slivers. From a kinematic point of view, this late cleavage is associated to thrust faults (Figure 5A) with top-to-the-south-west sense of shearing and strike-slip movements. Both sinistral (Figure 5B) and dextral movements have been measured.

Figure 5. Deformation associated with the serpentinite slivers in the Col de Nassirah and Kûfara areas. A, folding and shearing of sepentinites and schists, Shear zone: N155/20°NE, indicating a top to the WSW sense of shear; B, Serpentinites crosscutting the Eocene turbidites showing two main deformations,(i) a pervasive foliation (S1) (N40 and 75° dipping to the SE); and a (ii) discrete cleavage (S2) (N170 and 70 dipping to the east), S2 is parallel to the contact between serpentinites and Eocene turbidite; C, Tectonic contact between Eocene Flysch conglomerate and serpentinites, (contact: N170 dipping 75° to the east); D, Contact between basalts of the basement and serpentinites (N45/60°E); E, Lenses of serpentinized peridotites into serpentinites.

In the vicinity of col de Nassirah, towards the west, the contact between Eocene flysch and Mont Do massif is a N170 oriented strike-slip fault, with a slight inverse component (Figure 5C). Furthermore, in this area, the Eocene flysch shows kilometre size open folds with N-S axial planes (Figure 3A,B). These folds are compatible with reverse strike-slip faulting.

Kûfara area

On the western border of the Massif du Sud (Peridotite Nappe), several km-scale anastomosing serpentinite slivers are present and have thicknesses of 5–50 m (Figure 6). Basalts are commonly associated with the serpentinites as metre to kilometre-scale lenses in the basement (Figure 5D). The slivers follow an arcuate path whose concavity faces the east and whose chord is oriented N-S. They are steeply dipping to the east, and consist of foliated mylonitic serpentinites enclosing metre-size boudins of serpentinised peridotite (Figure 5E). Basement, Cretaceous-Paleocene and Eocene sedimentary rocks are folded with development of open, tight to isoclinal folds and schistosity parallel to the fold axial plane (Figure 6B). Folds and axial planes are overturned to the west. Axial planes of folds, schistosity in the sedimentary rocks, and foliation in serpentinite slivers are all parallel, suggesting that sedimentary rocks and serpentinite slivers share the same deformation. Serpentinite bodies and sedimentary rocks may be folded together (Figure 6C), a configuration that may explain branching and anastomosis of serpentinite slivers.

Figure 6. A, Geological map in the Kûfara area. Serpentinites crosscut both the Central Terrane and the Eocene turbidites; B, Simplified W-E cross-section showing the serpentinite slivers cutting across the basement, their orientation being parrallel to the main schistosity.

Structural synthesis

General features

Structural features observed in the study area are summarised in Figure 7. Serpentinites outcrop in two types of structural settings: (1) as shallow dipping basal sole of the main peridotite massifs, and (2) as slivers steeply dipping to the east, crosscutting the basement.

Figure 7. Schematic cross-section summarising the different field observations and the relationships of the serpentinite with the substrate.

In both structural settings, serpentinites share the same geometry and deformation styles. Both are deformed by a pervasive foliation, which is parallel to their main contact with the substrate. Serpentinite slivers cross cut all formations from the Paleozoic basement up to the Eocene Flysch. At the junction of the two types, the two foliations are associated, the deformation of the steeply dipping slivers being superimposed to the deformation of the sole although with a weaker intensity. Rocks of the substrate and serpentinite slivers share the same deformation orientations.

Kinematic method

All measures are reported as true (geographic) north (not magnetic north), the approximate mean magnetic declination in the centre of Grande Terre being of 12° to the east. Measured fault-slip data was plotted (lower hemisphere, equal-area projection) and analysed using Faultkin (v. 8.1) software (Marrett and Allmendinger 1990; Allmendinger et al. 2012) which are widely accepted and validated tools. The inversion technique of this software is based upon the contouring procedure of (Kamb 1959) and linked distribution statistic of Bingham (1974), and was used to calculate the maximum P (shortening), T (stretching), and B (intermediate) stress axes.

Kinematic results

Kinematic fault analysis has been done in several outcrops of the serpentinite slivers as well as on the sole of the Mont Do Massif, using slickenfibers, staircase geometry, and sigmoidal foliation or cleavage as shear sense indicators (Figure 8). In most of the slivers, lateral then thrust (with top-to-the-SW) senses of slip are dominant. Normal sense of slip is rare. In the basal sole of the Mont Do there is more variability with a greater proportion of normal motion with dip slip. This likely reflects the transition from a ductile (serpentinite) to a semi-brittle (serpentinised peridotite) material. Kinematic indicators are also present in the basement, though much less frequently, and are oriented consistently with those of the serpentinite slivers.

Figure 8. Kinematic data for several locations in the serpentinite slivers and the basal sole of the Mont Do Massif. In the stereonet P, T, and B are respectively the calculated direction of maximum stress axes of shortening, stretching, and intermediate. The coloured area are Kamb contours in standard deviations for P (2 shades of blue) and T (2 shades of red).

Superimposed deformations with a cleavage overprinting a pervasive foliation fabric have been observed in several places. Rather than being evidence for a several stages scenario these features likely reflect a more or less continuous process with the same orientation, confirmed by the fact that faults are all compatible in each outcrop. Kinematic indicators and their resulting tensors suggest a SW to W shortening direction.

AEM resistivity model

Method

The survey was carried out over the Boulouparis area, between the Mont Do Massif and the northwestern limit of the Massif du Sud (Figure 2); the SkyTEM system was used. The flight plan comprises a total of 197 linear kilometres and includes 11 lines oriented N 90 and 12 lines oriented N 105 (Figure 9), with a line spacing of 400 m. Along each line, an AEM data were acquired every ∼30 m.

Figure 9. Average resistivity soundings for each geological unit. Dashed lines represent the first and the third quartiles. Numbers of soundings: 1542 into Eocene Flysch; 596 into Koh basalts; 2001 into central terrane grauwacke and 916 into peridotite.

Generally speaking, AEM allows imaging the electrical conductivity/resistivity contrasts in the subsurface (Ward and Hohmann 1988). Injecting a current varying in time in a loop generates a transient magnetic field. This transient magnetic field excites the ground that emits a secondary magnetic field through the induction and diffusion of eddy currents. This latter is measured and inverted in order to estimate the electrical conductivity of the investigated rocks. The depth of investigation (DOI) of the method is variable and depends on the emitted magnetic moment, on the bandwidth used, on the subsurface conductivity/resistivity of the rocks and on the signal to noise ratio.

The SkyTEM system was developed for hydrogeophysical and environmental investigations by the HydroGeophysics Group at Aarhus University, Denmark (Sørensen et al. 2004; Auken, Christiansen et al., 2009). It is composed of (1) a transmitter coil which excites the subsurface, (2) a receiver coil for measuring the ground response; (3) a generator, as a power source; and (4) several navigation instruments such as GPS, tiltmetres, and laser altimeters in order to monitor the behaviour of the loop during the flight and locate the system in the space at all times. The system operates in a dual transmitter mode. The low moment, with a magnetic moment of approximately 3400 Am2 and time gates from about 7 µs to 500 µs for this survey, provides early time data for shallow imaging, and the high moment, with a magnetic moment reaching 156800 Am² and time gates from about 90 µs to 8.9  µs allows measuring later time data for deeper imaging.

AEM data are very sensitive to the ambient EM noise, especially in an anthropic environment. In order to improve the signal to noise ratio and reject unusable data from the dataset, several processing are applied to remove/reduce couplings with man-made installations and ambient noise from the signal. The processing scheme used is described in Reninger et al. (2020) and is based on the singular value decomposition (Reninger et al. 2011). Data were inverted using the Spatially Constrained Inversion algorithm (SCI) (Viezzoli et al. 2008): (1) each usable AEM data are translated into a 1D (AEM sounding) model divided into n layers, each defined by a thickness and a resistivity, and displaying the resistivity variations according to depth; during the inversion, constraints are applied vertically and spatially between nearby soundings (independently of flight lines). (2) The ground clearance of the transmitter is also inverted and the DOI is assessed as a final step in the inversion. Results were obtained running a smooth inversion (30 layers from 0 to 600 m deep), which is effective for imaging complex geological structures with the lowest dependency on the starting model but only displays a smoothed view of the subsurface. At this stage, each flight line can be displayed as a resistivity profile composed by all the associated AEM soundings.

A 2D model is then obtained by interpolating, on raster grids, the resistivity of layers falling into a depth or elevation range. This process is repeated in order to obtain different horizontal slices over the entire range of investigation.

Geological units resistivity

Figure 9 shows the average resistivity reading for each geological unit; dashed lines represent the first and the third quartiles for each formation. For this, we considered every AEM readings falling into each geological polygon of the geological maps. However, we didn’t consider the serpentinite polygon, since the width of several observed a serpentinite slivers is under the resolution capability of AEM, making statistics unreliable on this object. Moreover, serpentinite slivers can have a significant dip (from 80 to 40°), which is not compatible with a 1D representation of the subsurface (i.e. the average vertical resistivity readings). More generally, given the non-horizontal nature of the lithological contrasts, we truncated the resistivity sounding of each geological unit to 200 m, below, resistivity variations cannot be reasonably interpreted; only the peridotite nappe can be considered as sub horizontal environment.

The interpretation of the Figure 9 is not straightforward. However, Figure 9 gives an overview of the imaged resistivity for such a context and highlights some interesting trends that should be explored further. First, all the resistivity soundings show the same trend in the first ∼10 m, rocks are relatively conductive, then the resistivity increases suggesting a weathering profile. Second, the serpentinites under the peridotite nappe appear to be well differentiated from the others formations and is characterised by a resistivity between 20 and 50 ohm.m. Finally, under the weathering profile, all the formations are characterised by a resistivity higher than 200 ohm.m except the peridotite nappe, which appears more conductive with a resistivity between 100 and 300 ohm.m.

Figure 10 shows resistivity imaged between 15 and 20 m depth (Figure 9A), and a simplified geological map (Figure 9B). The depth range, just below the assumed weathering profile, highlights the main resistivity trends, while remaining consistent with the geological map and the observed serpentinites at the surface. As shown in Figure 9, the highest resistivity values are observed in the area where sedimentary rocks are dominant: schists, greywackes, and the Eocene flysch (> 250 Ω/m). Most of the lower resistivity values are clearly correlated to serpentinites (<50–100 Ω/m), either the flat lying serpentinite of the base of the Massif du Mont Do, especially in the northwest and around the Massif, or surprisingly the narrower and upright slivers in the centre and east. Basalts have a variable response, more difficult to define on a map. Peridotite of the Massif du Mont Do has values from <50 to 250 Ω/m, the lower resistivity values may be related to either the flat lying serpentinite, where the thickness of the peridotite is low, or to a possible weathered peridotite. The contrasting resistivity response of the serpentinite compared to other formations is a key for understanding the structure at depth of the serpentinite slivers and hence of the Peridotite Nappe. According to the observations, the presence of serpentinite should be related to a decrease in the imaged resistivity. Given the resolution limits of AEM, this decrease will be more or less marked depending on the thickness and the depth of the serpentinite and the background resistivity.

Figure 10. A, Resistivity grid between 15 and 20 m depth; B, simplified geological map of the surveyed area.

Resistivity profiles interpretation

Resistivity profiles (Figures 11 and 12) have been interpreted in the light of the surface geology. The western part of profile N° 1 (Figure 11A) provides an image of the relationships between the basement, here represented by the Boghen Terrane, and the Mont Do ultrabasic massif in the Oundamien-Aramuru area. The low resistivity values of the serpentinite sole of the Massif du Mont Do, form an eastward deepening continuous layer which is topped by a sheet of high resistivity values corresponding to the Boghen Terrane schists according to field data. What would be interpreted only on the basis of surface geology as a tectonic window of basement opened in the serpentinite sole of the Massif du Mont Do looks actually, on the basis of the AEM profile (Figure 11B), as a sheet of basement thrust westward upon serpentinite, thickening eastward, and rooted to the east. In section, the thrust surface shows long wave undulations (∼500 m), and the Aramuru crest made up of serpentinite appears to correspond to a reverse fault branching on the main thrust. To the west, in the basement, a thin undulating sheet of low resistivity value infers a sliver of serpentinite separating Central and Boghen terranes.

Figure 11. Resistivity profiles N°1 (A) and N°9 (C) and geological interpretations (B) and (D). Locations of sections on Figure 9.

Figure 12. Resistivity profiles 1-3-4-6-9 (locations on Figure 9) with structural interpretations. The tectonic contacts have been labelled A, B and C on the figure to make them easier to spot.

The western part of profile N° 9 shows a similar configuration (Figure 11C) with a western domain where the low resistivity values, typical of the serpentinite sole of the Massif du Mont Do, are plunging eastward below a zone of basement characterised by a high resistivity.

Therefore, it appears that the east of the Massif du Mont Do is an eastward dipping discontinuity of regional scale corresponding to a thrust of a compartment of basement and flysch cover over the ultrabasic Massif du Mont Do.

Figure 12 shows that the basement to the east of Mont Do, is cross cut by several low resistivity eastward dipping planes, some of them down to – 200 m depth, and generally pinching downward. They do correspond to the mapped serpentinite slivers (fault b and c), which are well-imaged by AEM owing to the resistivity contrasts.

We propose a geological interpretation of the resistivity profiles, with the major geological units, and structures (Figure 13).

Figure 13. Geological interpretations of AEM cross-sections (locations on Figure 9). The tectonic contacts have been labelled A, B and C on the figure to make them easier to spot.

Discussion

AEM data, structural analysis, and field data, provide relevant and complementary indications in the area of Boulouparis as to whether the Massif du Mont Do and the associated serpentinite slivers are related to the basement or to the Eocene Subduction-Obduction Complex.

So far, serpentinite elements have neither been found reworked in the Eocene Flysch (Maurizot et al. 2020c), nor in the basement greywackes, suggesting that Paleozoic ultramafic rocks, related to Kuah Ophiolite were never exposed before Eocene convergence. However, they could have been extruded after, owing to their particular rheological properties. In a general way, the patchy geometry of the basaltic occurrences of the Koh Unit, contrasts with the linear geometry of the serpentinite slivers. Serpentinite slivers are not preferentially associated to Koh Ophiolite unit. Actually, most of the serpentinite slivers observed are connected to the serpentinite sole of the Massif du Sud, to the east, suggesting a common origin. They are also connected to the sole of the Massif du Mont Do, to the west, linking the two massifs. In map view, the serpentinite slivers follow arcuate paths whose concavity faces the East, forming large concentric arc segments, roughly parallel to the base of the Massif du Sud, they are imaged by AEM and mapped in surface as slices dipping to the east, sharing the same structural pattern of overturned folds and thrusts, with a general westward vergence, a feature which is common to all Eocene subduction-obduction units in this area (Maurizot et al. 2020b).

In the study area, serpentinite slivers are more reasonably interpreted as slices pinched down from the overhanging serpentinite sole of the Peridotite Nappe, rather than ancient units extruded from the basement through younger formations. The Massif du Mont Do should be linked to the Massif du Sud.

In the online, updated multi-scale geological map of New Caledonia, This problem is clearly identified with areas where serpentinites are sometimes attached to the nappe or the Koh unit, without any arguments or data allowing to choose one of the 2 hypotheses. This problem is inherited from the different 1/50 000 geological maps, where the interpretation of these serpentinites differs from one author to another. Our study is not intended to be generalised to the whole New Caledonia, but to say that as recently published in the North (Gautier et al. 2016), there are strong arguments for linking these serpentinite slivers to the Peridotite Nappe.

The Peridotite Nappe emplacement (or obduction) is a typical example of continental subduction (Maurizot et al. 2020b). In this scenario an intra-oceanic subduction, is followed by the subduction attempt of a portion of continental lithosphere borne by the subducted plate, a situation that leads to the physically impossible underthrusting of light continental lithosphere under dense oceanic lithosphere and cause blocking of the subduction mechanism. In the case of the Peridotite Nappe of New Caledonia, the intra-oceanic NE dipping subduction nucleated during Late Paleocene in the oceanic south Loyalty Basin, located to the north of the continental Norfolk Ridge, and propagated southward until that ridge entered the subduction zone during Eocene. The consequent stalling of the subduction system was followed, after breaking off of the down going slab, by the exhumation of the previously subducted rocks, uplifting the deep oceanic mantle lithosphere of the south Loyalty Basin. The Peridotite Nappe thus represents the upper plate of the subduction system. All other units, from the autochthonous basement of the Norfolk Ridge and its cover, to the allochthonous nappes and terranes (Montagnes Blanches Nappe and Poya Terrane) that were pushed in front then stacked over the ridge, belong to the lower plate. This scenario accounts for the tectonic complexity of these last terranes compared to the relative simple structure of the Peridotite Nappe at large scale. It explains as well the general southwestern tectonic vergence imprinted in all overridden units. In this view, the basal serpentinite sole of the Peridotite Nappe which represents part of the mantle wedge (Iseppi et al. 2018) plays a major role. It has accommodated a large part of the deformation contrast between the weakly deformed overriding plate and the complexly folded and imbricated overridden plate.

The Peridotite Nappe is considered generally as a single sheet of mantle, latterly dismembered by post-obduction events (Paris 1981; Nicolas 1989) into several units and klippes of various morphologies. The nappe overlies diverse terranes, including basement terranes (this paper), Late Cretaceous to Eocene cover, Montagnes Blanches Nappe, and Poya Terrane (Maurizot et al. 2020c) (Figure 14). The relationships between the ultramafic terrane and these different substrates deserve consideration. Different structural settings may be distinguished (Figure 14A):

1. The Peridotite Nappe is absent in the northern HP-LT Metamorphic Belt, consistently with the exhumation, uplift, and unroofing process mentioned above.

2. In the low-grade schists of the metamorphic belt (Pic Ougne Zone), large elongated slivers of serpentinite and peridotite, mixed with ophiolitic mélange are interpreted as portions of the hanging wall of the subduction zone (mantle wedge) pinched down in the subduction zone (Maurizot et al. 1989; Gautier et al. 2016; Cluzel 2020), a situation somehow comparable to that of the Boulouparis area although being in a deeper structural level.

3. Southwards, the basement terranes of the backbone of Grande Terre (usually referred to as ‘Central Chain’ or axial range) are topped by kilometre-sized peridotite klippes (Central Chain Klippes), one of them being the Massif du Mont Do in Boulouparis area. The serpentinite sole of these klippes are characteristically complex and associated with steeply dipping serpentinite slices plunging within the basement. The Late Cretaceous to Eocene cover is there generally considerably thinned or absent as well as the Poya Terrane.

4. Larger klippes (5–10 km) are aligned along the southwestern coast (southwestern Coast Klippes). They dominantly overlie the Poya Terrane, although their northeastern flanks may locally rest directly over the Central Chain terranes.

5. Last but not least, the largest and most continuous area of Peridotite Nappe, the Massif du Sud, lies in the south of Grande Terre.

Figure 14. Conceptual model for the Peridotite Nappe structure in the surveyed area. A, Distribution and zonation of the Peridotite Nappe units. B, Structural sketch in Boulouparis area. C, Schematic SW-NE cross-sections.

The ‘Central Chain’ of Grande Terre is a major outcrop zone of basement terranes. It is suspected to be the remnant of a raised area inherited from a horst formed during the Late Cretaceous general rifting stage of Zealandia, separating two grabens or depocentres which are now the Diahot region in the NE and the SW coast – Nouméa area in the SW (Maurizot et al. 2020a). Over this salient, rugose, and rigid zone, tectonic erosion would be responsible during obduction for the scraping off of portions of the bottom of the overriding upper plate (Peridotite Nappe) by the downgoing lower plate and their pinching down as many serpentinite slivers in the main discontinuities of the basement.

A dextral oblique convergence model has been proposed by several authors for the emplacement of the subduction-obduction complex of New Caledonia (Cluzel et al. 2001; Gautier et al. 2016; Cluzel et al. 2021) on the base of diachronism and southward propagation of syntectonic deposits (Eocene Flysch) in front of the subduction system (Maurizot et al. 2020c) as well as on the patterns of fold and thrust faults, and kinematic indicators, in different geological units of Grande Terre (Cluzel et al. 2021). This model is consistent with our observations in the Boulouparis area, where almost all faults have a thrust or dextral components.

Despite a general flat-lying attitude of the Peridotite Nappe at large scale, in depth investigation by EM shows that some ultrabasic portions of the Peridotite Nappe are involved in the top-to-the-SW general tectonic vergence that took place during subduction-obduction process, culminating by the obduction of the Peridotite Nappe. The top-to-the-west overthrust of the basement formations on the northeastern flank of the Mont Do, may be suspected to occur in other units (e.g. Tene – Mé Adéo klippe). A model taking into account this tectonic style is proposed in Figure 14B. This model is rather similar to that proposed in Pic Ougne Zone (Gautier et al. 2016) except that it occurs in a higher structural level, in unmetamorphosed context, and without associated melange.

Conclusions

AEM proved to be a useful method to explore structure in a complex geological accretionary context, involving ultramafic units and convergent plates. This utility is based mainly on the low resistivity of the serpentinite (actually the lowest regionally), which constitutes a contrasting signature, easily recognisable within the ranges of resistivity of all lithologies. AEM is complementary to the magnetic method, which is commonly used for geological mapping. Magnetism characterises ultrabasic lithologies fairly well due to their high magnetite content, but the method does not allow geometries to be analysed as precisely as AEM. Magnetic signatures can be complicated by the presence of complex geometries acquired during structural phases, and the resolution of the magnetism decreases very rapidly with depth. Admittedly AEM costs more but offers direct 2D to 3D views of the subsurface, imaging rock slivers to several hundred metres depth and providing useful tectonic detail on the geometry of their contacts. Moreover, owing to their rheological properties, serpentinites and peridotites are crucial lithological units which, at orogen-scale, mark the most important interface between convergent plates. Serpentinite and peridotite, important rocks in subduction-obduction complexes, are especially amenable to thresholding by AEM.

The AEM survey coupled with outcrop-scale structural analysis, has allowed us to suggest a significantly different model to the one commonly proposed for Peridotite Nappe emplacement as a single homogeneous sheet. Some portions of the Peridotite Nappe are involved into major shear zones (thrust or strike-slip faults) with the basement and even if there is no important thickening by duplexing or internal thrust within the peridotite, these structures can be interpreted as being acquired during nappe emplacement, and acquired by shortening. Serpentinite plays a major role in this process by accommodating most of the deformation and by partitioning strain between the substrate and the nappe, i.e. between the lower and the upper plates. There may be lessons for other orogens from this new tectonic model of the New-Caledonian ophiolite.

In collision zones serpentinite are ubiquitous rocks and published methods proposed for differentiating the different types are rare. Recently, in New Caledonia, rare earth and trace element geochemistry has been successfully used by Raia et al. (2022) to distinguish subducted lower plate serpentinites from obducted upper plate serpentinites by their supra-subduction signature. We suggest for the future, that this type of geochemical approach could be tested in support of geophysics.

Acknowledgements

This work received the support of the BRGM, the funding agency Centre National de Recherche et Technologie Nickel et son Environnement (CNRT or National Centre for Research and Technology Nickel and its Environment, https://www.cnrt.nc) and the SGNC (Service géologique de Nouvelle-Calédonie) through OphioStruct project. Brice Sevin, Pierrick Rouillard, Julien Collot, Julie Jeanpert, Dominique Cluzel, Willy Foucher are thanked for their fruitful discussions in the lab and in the field. We also thank gratefully Nick Mortimer and an anonymous reviewer for their constructive remarks, which made it possible to greatly improve this article. We also thank Andy Nicol for the editorial handling.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article.