Magnetotelluric data from the Southeastern Capricorn Orogen, Western Australia: an example of widespread out-of-quadrant phase responses associated with strong 3-D resistivity contrasts

Magnetotelluric data from the Southeastern Capricorn Orogen, Western Australia: an example of... Summary Out-of-quadrant impedance phases (POQ) have been observed in several magnetotelluric (MT) data sets around the world, hampering the modelling and interpretation of the data. These anomalous responses are usually observed in small groups of sites and have been variously interpreted as due to electrical anisotropy, galvanic distortion, 2-D structures with large resistivity contrasts, or 3-D conductive bodies, such as L-shaped conductors. We present here what is possibly a unique land-based MT data set characterized by an exceptionally large number of sites with POQ. The MT data were collected in the southeastern margin of the Capricorn Orogen (Western Australia). This is an area with a complex geological history, which is evident in the MT responses. Given the characteristics of the study area, which includes very high resistivity Archean terrains surrounded by low resistivity Palaeoproterozoic basins, strong resistivity contrasts involving complex 3-D structures are a possible cause of this unusual behaviour. This is confirmed by 3-D forward modelling using models based on the general geology of the study area, as well as by the 3-D inversion results. A comparison of the real and synthetic data, derived from 3-D forward modelling, using the same parameters suggests that the proposed scenario is a realistic explanation for most of the anomalous phases observed in this data set. Synthetic data are not affected by galvanic distortion or electrical anisotropy, so the good match observed between the real and synthetic MT responses is likely due to 3-D inductive effects. Previous explanations for POQ responses, such as blocks with electrical anisotropy and 3-D L-shaped conductors, are not required and are considered to be less likely causes of the POQ responses on geological grounds. Australia, Magnetotellurics, Cratons 1 INTRODUCTION A large multi-disciplinary geoscience project ‘Distal Footprints of Giant Ore Systems’ is being undertaken in the Capricorn Orogen, in the central part of Western Australia. The project involves large scale geophysical data acquisition including passive seismic recordings, airborne electromagnetics and a major magnetotelluric (MT) study (Aitken et al. 2015). A key objective of the deep geophysical component of the project is to map the boundaries of major Archean basement blocks beneath the cover of younger sedimentary rocks, because these are considered an important indicator of regional prospectivity for several important kinds of mineral deposits (McCuaig & Hronsky 2014). An unusual characteristic of the MT data set from the Capricorn Orogen is the large number of sites with phases out-of-quadrant (POQ) at long periods. The phases for the off-diagonal components of the impedance tensor will normally lie in the first (0° < phs. <90°) and third quadrant (–90° < phs. < –180°) for the XY and YX components, respectively. The authors are not aware of such a large number of sites with POQ (56 sites of a total of 83 measured across an area of about 40 000 km2) observed elsewhere in the world, possibly making this a exceptional MT data set. Although uncommon, POQ MT responses have been observed in other data sets collected elsewhere in the world and several studies have attempted to explain the origin of this phenomenon. POQ has been explained by the presence of electrical anisotropy (e.g. Heise & Pous 2003; Pek 2009), galvanic distortion (e.g. Chouteau & Tournerie 2000; Lilley & Weaver 2010), 2-D structures with large resistivity contrasts (e.g. Selway et al. 2012), or 3-D conductive bodies (e.g. Lezaeta & Haak 2003; Ichihara & Mogi 2009; Thiel et al. 2009; Ichihara et al. 2013). Anomalous phases due to electrical anisotropy appear mainly in sites located above an anisotropic block and near its borders (e.g. Heise & Pous 2003). 3-D models proposed to explain POQ generally include conductive structures with contact angles of 90°, such as L-shaped bodies (Ichihara & Mogi 2009) or a hollow square (e.g. Pous et al. 2002; Weckmann et al. 2003). In the L-shaped bodies case, the locations of the POQ responses are restricted to the corner where the two conductors that make up the L-shape are connected, whilst in the hollow square case POQ occur in recordings directly above the conductive body, but not inside or outside of the body. Thus, there are clear restrictions on the locations of the sites with anomalous phases relative to the conductive body that is causing the POQ. Such local-scale explanations are very unlikely to be applicable to the Capricorn Orogen data set because of the large number of sites involved and the fact that these occur over an area of around 200 km × 200 km. Seeking an alternative explanation, we have attempted to explain this phenomenon using 3-D forward modelling of data from the southeastern part of the Carpicorn Orogen (Fig. 1). Figure 1. View largeDownload slide (a) Location of the study area. (b) Location of the sites with anomalous YX phases (red triangles) independent of observation angle. (c) Location of the sites with anomalous YX phases (red triangles) for a specific observation angle, rotation of the impedance tensor equal to zero (see text for more information). Figure 1. View largeDownload slide (a) Location of the study area. (b) Location of the sites with anomalous YX phases (red triangles) independent of observation angle. (c) Location of the sites with anomalous YX phases (red triangles) for a specific observation angle, rotation of the impedance tensor equal to zero (see text for more information). 2 GEOLOGICAL SETTING The Proterozoic Capricorn Orogen is a zone of deformed and metamorphosed igneous and sedimentary rocks located in between the Archaean Pilbara and Yilgarn Cratons (Fig. 1a). The Orogen comprises the deformed margins of the two Cratons, the Archaean–Palaeoproterozoic Gascoyne Province and a series of Palaeoproterozoic–Mesoproterozoic sedimentary basins, for example the Collier, Yerrida and Bryah Basins (Fig. 1). There is an inlier of Yilgarn Craton granitoid-greenstone rocks, the Marymia Inlier, immediately to the north of the Yilgarn Craton. The Capricorn Orogen has a very long and complex tectono-magmatic history related to the collision of several continental basement terranes and also subsequent intraplate deformation events, as described by Sheppard et al. (2016). 3 MT DATA SET A total of 83 broad-band magnetotelluric stations (BBMT) have been recorded to date in the southeast part of the Capricorn Orogen (mostly within the Bryah-Padbury and Yerrida Basins) and adjacent Yilgarn Craton (Fig. 1). This data set will be referred herein as the SECO-data set. The field area is remote and difficult to access. Logistical constraints required the data to be collected as traverses along roads and tracks. The MT data acquisition was undertaken in three stages. The easternmost sites were recorded in May 2011 (Dentith et al. 2014), while the westernmost sites were acquired in October 2013. Twenty new BBMT sites were collected in October 2015, mostly in the Collier Basin around the Marymia Inlier. Data were acquired using Metronix ADU07e and Phoenix MTU-5A recorders with induction coils, and non-polarized electrodes (Pb/PbCl2) with a typical dipole length of 100 m. The x-axis was oriented in the magnetic N–S direction, with the positive direction pointing to the north, and the y-axis in the E–W direction, positive to the east. The data were recorded at each location for an average of 48 hr, and the typical spacing between stations is 10 km. Archean rocks in the upper crust are expected to have high resistivities (e.g. Jone 1999; Selway et al. 2009), whilst the Proterozoic basins are expected to have lower resistivity values, due to the presence of shales and iron-rich formations. These strong resistivity contrasts and the complicated outcrop distribution of basins and basement, resulting from the area's complex geological history, are reflected in the complexity of the MT data. 3.1 Phases out of quadrant An important characteristic of the SECO-data set is the large number of sites with phases that are out of quadrant (above 90°) at long periods (Fig. S1). This behaviour is typically observed at periods longer than 10 s and across the whole study area, and there is no obvious relationship between the location of the sites with anomalous phases and the outcropping geology (Figs 1 and S1). This effect is strongly dependent on the rotation angle of the impedance tensor (Z). Z is a complex tensor which relates the orthogonal electric and magnetic fields recorded at each MT site. Thus, it contains information about dimensionality and direction (Simpson & Bahr 2005). A mathematical rotation of the impedance tensor from its measuring axis will provide information about the electrical dimensionality of the data, but in our particular case, it also highlights a strong dependence of the occurrence of anomalous phases on this rotation angle. Data were recorded with the x-axis in the magnetic N–S direction. This corresponds to a rotation angle of 0°, whilst a rotation angle of 90° will lead to the x-axis being parallel to the E–W direction. This rotation modifies all components of Z and some of the components can become zero (e.g. 2-D structures where one of the axes is aligned along the electrical strike, in which case the off-diagonal components of Z are zero). However, regardless of dimensionality (1-D, 2-D, 3-D structures) and rotation angle, phases are not expected to exceed the ordinal quadrant limits. Fig. 1(b) shows the location of the sites with YX-POQ (in red) independent of rotation angle, that is those sites for which the YX phases are out of quadrant for any rotation angle between 0° and 180°. Fig. 1(c) shows the location of the sites with YX-POQ for a rotation angle for the impedance tensor equal to zero. Note the large number of distorted sites located in, and around, the Marymia Inlier. We concentrate on stations from this area (grey box) in our analysis of the POQ. Fig. 2 shows the dependence of the phase response on the impedance rotation angle using the YX phases for site MRY022 (yellow circle in Fig. 2b). The purple region in Fig. 2(a) shows the periods in which the phases are outside the third quadrant (Y-axis) for the different rotation angles (X-axis). The black and red vertical lines represent the limits of this out of quadrant range of angles, for the specific period of 10 s (0.1 Hz, white dashed line). This is in general the lowest period where POQ occurs across the study area. A clockwise rotation from the black line keeps the phases within the ordinal quadrant. The same effect is observed with an anticlockwise rotation from the red line. Figure 2. View largeDownload slide (a) YX phases for site MRY022 according to the observation angle. White dashed line indicates a period of 10 s. Black and red lines correspond to the limits of the area where the phases are within the third quadrant, for the period chosen. (b) Range of angles for which the YX phases are within the ordinal quadrant (in grey), for a period of 10 s. Yellow circle: Site MRY022. Surface geology is in the background. (c) Orientation of the electric field (blue lines) for site MRY022 when the phases are out of quadrant, for a period of 10 s. Red and black lines: range of angles for which the YX phases are out the third quadrant. Black dashed line: Trend of the Marymia Inlier limits. Figure 2. View largeDownload slide (a) YX phases for site MRY022 according to the observation angle. White dashed line indicates a period of 10 s. Black and red lines correspond to the limits of the area where the phases are within the third quadrant, for the period chosen. (b) Range of angles for which the YX phases are within the ordinal quadrant (in grey), for a period of 10 s. Yellow circle: Site MRY022. Surface geology is in the background. (c) Orientation of the electric field (blue lines) for site MRY022 when the phases are out of quadrant, for a period of 10 s. Red and black lines: range of angles for which the YX phases are out the third quadrant. Black dashed line: Trend of the Marymia Inlier limits. Fig. 2(b) shows the angles for which the YX phases are within and out of the third quadrant, for each site for a period of 10 s. Most of the sites with anomalous phases at this period are located in the northern part of the study area. 4 DATA ANALYSIS MT tensor analysis was performed on the SECO-data set using several different techniques, in order to understand the dimensionality of the regional electrical variations, as well as try to identify galvanic and inductive causes of the anomalous phases, since this anomalous behaviour has been associated with both causes (see Section 1). The analysis reveals clear 3-D geoelectrical behaviour and extreme complexity for most of the sites recorded in this area. Given this, several 3-D electrically isotropic synthetic models were created to try and reproduce the anomalous phases, since 3-D complex structures involving strong resistivity contrasts are considered as a potential cause of the POQ. Comparison of the observed and synthetic MT responses from those sites with anomalous phases helps to understand the mechanisms related to this unusual behaviour since, for example, the synthetic data are not affected by galvanic distortion. In this section, we show the results of the data analysis for both observed and synthetic responses, while their comparison is addressed in the discussion. 4.1 Observed data 4.1.1 Phase tensors Fig. 3 shows the phase tensor (Caldwell et al. 2004), computed using the MTpy package (Krieger & Peacock 2014), for each station for each of six different periods from 0.01 to 600 s. The phase tensor is one of the most common techniques to analyse MT data, since it is not affected by galvanic distortion and does not make any assumptions about the dimensionality of the regional electrical variations (Caldwell et al. 2004). Its graphical representation is as an ellipse, where both axes are equal (circle) in a 1-D medium, whilst in a 2-D case the axes are parallel or perpendicular to the strike of the structure. For 3-D structure the phase tensor is non-symmetric and the skew is the angle which gives the rotation of the major axis away from an identically shaped ellipse (Chave & Jones 2012). Skew values larger than ±5 are usually considered to be an indication of a 3-D resistivity distribution. Figure 3. View largeDownload slide Phase tensor ellipses for each BBMT site at six different periods. Figure 3. View largeDownload slide Phase tensor ellipses for each BBMT site at six different periods. The results of this analysis show the obvious three-dimensionality of the data, with skew values reaching greater than ±10 at periods greater than 1 s, while at shorter periods, skew values are usually lower than ±5. One of the most intriguing aspects revealed by Fig. 3 is the behaviour of the data at sites located in the vicinity of the Marymia Inlier (grey box in Fig. 1c). At 1s, there is a group of five sites with a skew lower than –10 and the major axis of the phase tensor aligned in a NW–SE direction. At 10s, some neighbouring sites are included in this group, ten in total, but in this case the major axis of the phase tensor has a slight clockwise rotation relative to the 1 s data. At 100 s the group comprises thirteen sites, all of them with similar high skew values. The major axis of the phase tensor shows a further clockwise rotation, being now in a NNE–SSW direction. Finally, at 600 s this group of sites with similar behaviour is no longer apparent, even though elsewhere in the area the responses at 100 and 600 s are quite similar. 4.1.2 WAL invariants An analysis of the WAL invariants (Weaver et al. 2000) was performed using the WALDIM code (Martí et al. 2009). This method also provides information about the dimensionality and the orientation of the electrical structure. As observed using the phase tensors, many of the data behave three-dimensionally at periods longer than 1 s (not shown here). There is a set of eight rotational invariants known as WAL invariants, seven independent (I1 to I7) and one dependent (Q). Invariants I3 to I7 and Q can be used to determine the type of dimensionality and identify galvanic distortion (Martí et al. 2009). Invariants I7 and Q provide information about the dimensionality, since I7 is non-zero only for 3-D cases and Q is zero only for 1-D. Invariants I5, I6 and I7 are related to galvanic distortion, while I3 and I4 have information about the degree of 2-D anisotropy (Chave & Jones 2012). Fig. 4 shows the values of the invariants I3 to I7 for two sites from the SECO-data set. Fig. 4(a) shows the WAL invariant of one site with POQ (see Fig. 1c for location), while Fig. 4(b) shows the same invariants for one of the sites with phases always within their respective quadrants regardless of the rotation angle (see Fig. 1c for location). Both sites show non-zero values of the invariants I3, I4 and I7 for long periods, indicating the 3-dimensionality of the data. Figure 4. View largeDownload slide WAL invariants I3, I4, I5, I6 and I7. (a) Example of a site with anomalous YX phases. (b) Example of a site with YX phases in the ordinal quadrant (see Fig. 1c for location, open circles). Figure 4. View largeDownload slide WAL invariants I3, I4, I5, I6 and I7. (a) Example of a site with anomalous YX phases. (b) Example of a site with YX phases in the ordinal quadrant (see Fig. 1c for location, open circles). The main difference between the sites with and without POQ is in the invariants I5 and I6. Sites with phases within the ordinal quadrant (Fig. 4b) have relatively low values for the invariant I6 for the whole period range, while sites with POQ (Fig. 4a) show higher values at long periods. Meanwhile, invariant I5 is very small at high frequencies for both types of site, as well as at long periods for those sites with phases within the ordinal quadrant. For those sites with POQ the value of this invariant at long periods is clearly non-zero. 4.1.3 Groom and Bailey decomposition The Groom and Bailey decomposition technique is used to separate local and regional effects in the MT responses, and resolve the geoelectrical strike. There are four independent parameters in the distortion tensor, including three tensor suboperators: twist, shear and anisotropy, and one scalar: site gain, which is the only indeterminable parameter. Unlike the phase tensor and WAL invariants, which make no assumptions about the dimensionality, this technique assumes that the Earth has a 2-D regional conductive structure. Therefore, large 3-D inductive effects are not accounted for directly in the method. Nonetheless, this analysis can provide some information about the presence of 3-D induction, since the twist and shear parameters will show frequency dependence under these circumstances (Simpson & Bahr 2005). Fig. 5 shows the unconstrained Groom & Bailey (1989, 1991) galvanic distortion parameters twist and shear for the sites with POQ (Fig. 5a), and without POQ (Fig. 5b). There are clear differences between these two groups of data. Sites with POQ show very high values of shear, rising quickly to its physical limit, ±45°. Such values are normally regarded as severe current channelling effects, for example Jones et al. (1993). For this group of sites, the twist shows an approximate frequency-independent behaviour at short periods with a rapid increase in its value at periods greater than 1 s. In contrast, sites with phases within the ordinal quadrant show less extreme behaviour (Fig. 5b). In general, the shear is lower than 30° and never reaches the limit of ±45°. The twist is scattered at periods longer than 10 s, but it does not show the steep slopes observed in the sites with POQ (Fig. 5a). Figure 5. View largeDownload slide Groom and Bailey parameters for the sites located in the Marymia Inlier area (see Fig. 1c for location, grey box). (a) Sites with YX anomalous phases. (b) Sites with YX phases within the third quadrant. Figure 5. View largeDownload slide Groom and Bailey parameters for the sites located in the Marymia Inlier area (see Fig. 1c for location, grey box). (a) Sites with YX anomalous phases. (b) Sites with YX phases within the third quadrant. 4.2 Forward modelling Given the unusual number of sites with POQ that are located across a large area, the distinctive 3-D geoelectrical behaviour demonstrated by the various data analysis methods described above, and the expected strong resistivity contrasts between the Archean units and the Palaeoproterozoic basins, we consider 3-D variations involving strong resistivity contrasts as a possible cause of the anomalous phases. In order to analyse this possibility, several 3-D electrically isotropic synthetic models were created to try and reproduce the observed POQ. Due to the variety and complexity in the responses observed in the SECO-data set, we focus the analysis on those data from the vicinity of the Marymia Inlier (grey box in Fig. 1c). Forward modelling was carried out using the ModEM code (Egbert & Kelbert 2012; Kelbert et al. 2014). The responses have been calculated for the full impedance tensor for 18 periods, in 61 sites located according to the actual MT sites. All the models have the same basic structure, which consists of Archean units with resistivities of 10 000 Ωm and basins with resistivities of 50 Ωm. The basins are assumed to be 8.4 km deep, with Archean basement beneath, also with a resistivity of 10000 Ωm (Fig. 6). This overall structure is derived from the surface geology, while the resistivity values are assigned taking into account the actual MT data. Variants on this simplest model (Model S) were created mainly by adding conductive bodies at different depths and in different areas, as well as reducing the thickness of the youngest Palaeoproterozoic basin (the Yerrida basin). Fig. S2 shows some of the models tested, as well as the responses predicted in comparison to the observed MT data. Figure 6. View largeDownload slide Example of two 3-D synthetic models used to try and explain the observed POQ (more models are shown in Fig. S2). Model S: General background model for 0–8.4 km depth; at greater depths the model is a homogeneous half-space of 10 000 Ωm. Model S-7: Same shallow structure than Model S, but with an E–W trending conductive body (2 Ωm) eastward the Marymia Inlier at depth (8.4– 9.8 km). Red triangles, sites with XY anomalous phases; Yellow triangles, sites with YX anomalous phases. Figure 6. View largeDownload slide Example of two 3-D synthetic models used to try and explain the observed POQ (more models are shown in Fig. S2). Model S: General background model for 0–8.4 km depth; at greater depths the model is a homogeneous half-space of 10 000 Ωm. Model S-7: Same shallow structure than Model S, but with an E–W trending conductive body (2 Ωm) eastward the Marymia Inlier at depth (8.4– 9.8 km). Red triangles, sites with XY anomalous phases; Yellow triangles, sites with YX anomalous phases. Several models, including the simplest one (Model S, Fig. 6), lead to phases being out of quadrant; but only those models including a conductive body eastward of the Marymia Inlier induce the YX phases (the main component affected in the real data set) to leave the third quadrant (e.g. Model S-7, Fig. 6). This demonstrates that strong 3-D resistivity contrasts are possibly associated with the POQ observed in complex geological contexts, such as our study area. 4.2.1 Synthetic MT responses Fig. 7 shows the phase tensor, WAL invariants and Groom and Bailey parameters calculated from the responses of the synthetic models S and S-7 (Fig. 6). This corresponds to those sites where POQ has been reproduced, such as site MRY22 (Figs 7a and c). As observed in the SECO-data set, the phase response is also strongly dependent on the observation angle (compare Fig. 7a with Fig. 2a). Figure 7. View largeDownload slide Synthetic model responses. (a) YX phases for the site MRY022 (response from synthetic Model S-7) according to the observation angle. (b) Phase tensor ellipses at 1s period for the sites located on and around the Marymia Inlier (data from synthetic Model S). (c) WAL invariants I3, I4, I5, I6 and I7 for the site MRY022 (response from synthetic Model S-7). (d) Groom and Bailey parameters for the three sites with YX anomalous phases obtained in the Model S-7 (see Fig. 6 for location). Figure 7. View largeDownload slide Synthetic model responses. (a) YX phases for the site MRY022 (response from synthetic Model S-7) according to the observation angle. (b) Phase tensor ellipses at 1s period for the sites located on and around the Marymia Inlier (data from synthetic Model S). (c) WAL invariants I3, I4, I5, I6 and I7 for the site MRY022 (response from synthetic Model S-7). (d) Groom and Bailey parameters for the three sites with YX anomalous phases obtained in the Model S-7 (see Fig. 6 for location). Fig. 7(b) shows the phase tensor at 1 s for Model S in the vicinity of the Marymia Inlier. Here, the SECO-data set shows a clear orientation of the major axis of the phase tensor, perpendicular to the trend of the inlier (Fig. 3). At 1s the major axis of the phase tensor in those sites located on the inlier has the same orientation as observed in the real data, although we have not been able to reproduce the same clockwise rotation with increasing period observed in the SECO-data set. The SECO-data set shows a clear difference in the WAL invariants between the sites with phases in the ordinal quadrant and those sites with POQ (Fig. 4). The responses from the synthetic models show a similar behaviour for this last group of sites: non-zero values for all the invariants represented (I3 to I7) at long periods (compare Fig. 7c with Fig. 4a). Note the good correlation, in both values and trend, of all the invariants at long periods. Fig. 7(d) shows the unconstrained twist and shear angles of the Groom and Bailey decomposition for Model S-7 for those sites with POQ (see Fig. 6 for location). As observed in the SECO-data set (Fig. 5a) the shear rises quickly to 45°, while the twist shows an approximate frequency-independent behaviour at short periods with a rapid increase in its value at periods longer that 1s. 5 DISCUSSION The most remarkable characteristic of the SECO-data set is the large number of sites with anomalous phases. This unusual behaviour has been related to electrical anisotropy (e.g. Heise & Pous 2003; Pek 2009), galvanic distortion (e.g. Chouteau & Tournerie 2000; Lilley & Weaver 2010), 2-D structures with large resistivity contrasts (e.g. Selway et al. 2012), or 3-D conductive bodies (e.g. Lezaeta & Haak 2003; Ichihara & Mogi 2009; Thiel et al. 2009; Ichihara et al. 2013). Among these possibilities, we consider the 3-D structures with strong resistivity contrasts as the most likely cause of the POQ. This assumption is based on the geological characteristics of the study area, with Archean units surrounded by Palaeoproterozoic basins (e.g. Marymia Inlier, see Fig. 1) resulting in complex 3-D structures with potentially strong resistivity contrasts, as well as the results obtained from the data analysis, which reveals the area to be 3-D from the geoelectrical point of view. Other significant characteristic of the SECO-data set is that the anomalous phases show a strong sensitivity to the rotation angle of the impedance tensor (Figs 1 and 2). There is a range of angles for which the phases remain within their respective quadrants. Such behaviour has been observed in other studies, for example Jones et al. (1993) and Heise & Pous (2003), corresponding to those angles with the direction of the structural strike, in the case of 2-D isotropic models, or with the anisotropy strike, in the case of 2-D anisotropic models. A more detailed analysis of Fig. 2(b) in the Marymia Inlier area, highlights a correlation between the range of angles for which the phases are within the ordinal quadrant and the average trend of the inlier itself. Additionally, Fig. 2(c) shows the orientation of the electric field for the rotation angles of the impedance tensor for which the YX phases are out of the third quadrant. There is also a clear relationship between the trend of the inlier and the electric field orientation for those observation angles. Thus, YX phases are running out of quadrant when the electric field is approximately parallel to the inlier. All this can be interpreted as evidence that the strong resistivity contrasts expected along the contacts between the Archean basement and Palaeoproterozoic basins play an important role in the distortion that causes the phases to run out of quadrant. A similar behaviour has been observed in 2-D structures with a strong resistivity contrast between the basement and the cover (e.g. Selway et al. 2012). Thus, similar mechanisms could be involved in the responses observed here, although a 3-D approach is required due to both the characteristics of the MT data observed and the geological setting. This idea of strong 3-D resistivity contrast as main cause of the POQ is supported by the good agreement observed between the MT responses of the observed and synthetic data. Firstly, the orientation of the phase tensors at periods greater than 1 s (Fig. 3) could be largely due to the characteristics of the contact between the Marymia Inlier and the Palaeoproterozoic basins, as demonstrated by the synthetic model responses (Fig. 7b). Secondly, extremely high values of shear (Figs 4 and 7d) are typically indicative of severe current channelling effects at these periods (e.g. Jones et al. 1993). As the ± 45° limit is approached, the electrical field is basically propagating in one direction (Chave & Jones 2012). Both the real and synthetic data sets are characterised by a rapid rise in the shear values, reaching ±45° at long periods. The synthetic data are not affected by galvanic distortion, thus the strong period dependence of the shear and twist at periods between 1 and 100 s (Fig. 7D) indicates there are inductive effects associated with the E–W conductive body included in Model S-7 (e.g. Ledo et al. 2002). The similarity in the decomposition parameters between the synthetic data and the SECO-data set is an indication that both could be distorted by similar mechanisms. Therefore, 3-D inductive effects would be directly related to these anomalous phases, thus discarding the galvanic distortion as a potential cause. Finally, the dimensionality criteria according to the WAL invariants values, suggests a three-dimensionality affected, or not, by galvanic distortion for both real and synthetic data sets. Lilley & Weaver (2010) used the rotational invariants to analyse the anomalous phases in a case of galvanic distortion of a 2-D structure. Evidence for the anomalous phases observed in the SECO-data set being the result of galvanic distortion cannot be inferred from the findings presented here. WAL invariants from synthetic data with POQ (Fig. 7c) show a comparable distribution to that observed in the real data (Fig. 4a). Given this similarity, both data sets could be distorted by similar mechanisms, which do not include galvanic distortion or electrical anisotropy, both absent in the synthetic data. In summary, all the parameters analysed point to inductive effects involving 3-D structures as the primary cause of most of the POQ recorded in the SECO-data set. This effect arises due to the presence of various 3-D conductive bodies in combination with the strong resistivity contrast expected at the interface between resistive Archean rocks and the less resistive rocks of the Palaeoproterozoic basins. 5.1 3-D inversion model Taking into account the results obtained from the data analysis, 3-D inversion techniques were considered appropriate tools to perform the modelling of the SECO-data set. Furthermore, the interpretation of an inductive cause for the POQ phenomenon would be supported if the inversion models can achieve a good fit to the observed data with anomalous phases. 3-D inversion models were computed with the ModEM code (Egbert & Kelbert 2012: Kelbert et al. 2014). No rotation of the impedance tensor was applied to the data for the inversion, which used the four components of the impedance tensor. Initially, only 35 sites with phases in the ordinal quadrants for the observation angle were included in the inversion process, for 12 periods and covering the whole period range (stage 1). Subsequently, sites with anomalous phases were added to the inversion (58 sites in total), but only for those periods unaffected by the anomalous phases (T < 1 s) (stage 2). Finally, those periods for which the phases run out-of-quadrant were included in the inversion (stage 3). Fig. 8 shows map views of the 3-D inversion model at different depth for the three stages of inversion mentioned above. Comparison of the models allows recognition of aspects of the model required to fit the anomalous phases. The main differences are observed at depth, where the resistivity features labelled HR and LR appear better defined and with more lateral continuity. This is consistent with the synthetic models performed to reproduce POQ, where the anomalous phases are mainly due to the strong resistivity contrast expected between the Archean units and the Palaeoproterozoic basins (Figs 6 and S2 in SM). New sites with POQ appear by adding conductive bodies in the basins or modifying the limits of the Archean (Fig. S2). Thus, both inversion and synthetic models are coherent regarding the main characteristics (e.g. resistivity value, lateral extension) of the main structures related to the POQ. Figure 8. View largeDownload slide Map views of the 3-D inversion models at different depths. Stage 1: Inversion performed with 35 sites for the whole period range; no POQ are included in the inversion process. Stage 2: Inversion performed with 58 sites for periods shorter than 1 s; no POQ are included in the inversion process. Stage 3: Inversion performed with 58 sites for the whole period range; POQ are included in the inversion process. Figure 8. View largeDownload slide Map views of the 3-D inversion models at different depths. Stage 1: Inversion performed with 35 sites for the whole period range; no POQ are included in the inversion process. Stage 2: Inversion performed with 58 sites for periods shorter than 1 s; no POQ are included in the inversion process. Stage 3: Inversion performed with 58 sites for the whole period range; POQ are included in the inversion process. A good data fit has been achieved in the inversion performed with 58 sites for the whole period range, even for those sites with anomalous phases. Fig. S1 shows the comparison between the observed and predicted responses (stage 3 model) for some sites with POQ, for the four components of the impedance tensor. Thus, the results obtained from the inversion models support the idea of strong 3-D resistivity contrasts as a main cause of the anomalous phases. 5.2 Electrical anisotropy Given the geological complexity of the study area we cannot ignore the possibility of there being electrical anisotropy affecting the SECO-data set. To explain the POQ as a result of electrical anisotropy an outcropping anisotropic block with anisotropic azimuth different to the structural azimuth is required (e.g. Heise & Pous 2003; Pek 2009). The model proposed by Heise & Pous (2003) consists of an anisotropic block underlain by an anisotropic layer with anisotropy strikes perpendicular to each other, whilst in the model proposed by Pek (2009), only the anisotropic block is needed (Marti 2014). These models require an outcropping anisotropic block and all the sites with anomalous phases must be located above it. Outside the block, the distortion effects do not cause the phases to be out of quadrant. In view of these requirements, the large number and location of the sites with POQ (Fig. 1), and the good agreement between the responses of the SECO-data set and the synthetic models; we consider it unlikely that electrical anisotropy is the primary cause of the POQ. 6 CONCLUSIONS We present here what is possibly a unique land-based MT data set with a remarkable number of recording sites with phases out of quadrant. This effect is observed mainly in the YX component and extends across the whole study area, an area of around 40 000 km2. Here we summarize the main findings: Dependence on the observation angle: One of the most striking characteristics of this data set is the dependence observed between the occurrence of POQ and the rotation angle of the impedance tensor. The main direction of the structures that generate strong resistivity contrasts seems to be controlling the range of angles for which the phases are within the ordinal quadrant. Therefore, the phases come into their respective quadrants as the observation angle approaches this direction. Thus, the emergence of anomalous phases is strongly dependent on the relationship of both, the observation angle (Z rotation angle) and the trend of the structure that generates the strong resistivity contrast. 3-D inductive effects as a main cause of POQ. This is based on: The 3-D forward models and the good agreement between the MT responses from both real and synthetic data, show that the observed POQ are most likely caused by strong resistivity contrasts with complex 3-D geometries. 3-D inversions achieve a good data fit, even for the anomalous phases. This is a key result in the inversion of this particular data set, which can be interpreted as indicative of the inductive nature of the POQ phenomenon in this area. Geometry of the 3-D structures: It is important to note that 3-D structures with angles of exactly 90 degrees are absent in the synthetic models here presented. Previous work reproducing POQ with 3-D models include conductive bodies with 90 degree angles, such as L-Shaped bodies or hollow squares (e.g. Pous et al. 2002; Weckmann et al. 2003; Ichihara & Mogi 2009). This opens up numerous possibilities to explain anomalous phases as caused by complex 3-D structures. Electrical anisotropy: The presence of electrical anisotropy remains possible but we do not consider that this is the primary cause of such a large number of sites with anomalous phases. If some sites with POQ are recorded in areas with anisotropy, the anisotropic azimuth would probably be included in the range of angles for which the phases are in the ordinal quadrant (Fig. 2b). POQ appear when the observation angle is different from both the direction of the structures that generate strong resistivity contrasts (as shown in this study) and the anisotropic strike (e.g. Pek 2009; Heise & Pous 2003). This needs to be addressed in future work, in order to better understand the MT responses observed in this data set, since in some sites the anomalous phases are likely due to a combination of effects. Acknowledgements This work was supported by resources provided by the Pawsey Supercomputing Centre with funding from the Australia Government and Government of Western Australia. Authors thank Gary Egbert and Anna Kelbert for providing the ModEM code, as well as Naser Meqbel for providing the 3D-Grid visualizer. Moombarriga Geoscience collected the data described here in some cases with the assistance of personnel from the Geological Survey of Western Australia and the Centre for Exploration Targeting. The assistance of Ray Addenbrooke deserves particular acknowledgement. We also thank the various stations owners and the Plutonic Gold Mine for access to the field area. The editor, Ute Weckmann, and the reviewers, Kate Selway and Stephan Thiel, are thanked for the useful comments, which helped to improve the manuscript. REFERENCES Aitken A.et al.  , 2015. A major geophysical experiment in the Capricorn Orogeny, Western Australia, in Proceedings of the ASEG Extended Abstracts 2015: 24th International Geophysical Conference and Exhibition , pp. 1– 5, Australian Society of Exploration Geophysicists. Google Scholar CrossRef Search ADS   Caldwell T.G., Bibby H.M., Brown C., 2004. The magnetotelluric phase tensor, Geophys. J. Int. , 158( 2), 457– 469. https://doi.org/10.1111/j.1365-246X.2004.02281.x Google Scholar CrossRef Search ADS   Chave A.D., Jones A.G., 2012. The Magnetotelluric Method , pp. 1–584, eds Chave A.D., Jones A.G., Cambridge University Press. Google Scholar CrossRef Search ADS   Chouteau M., Tournerie B., 2000. Analysis of magnetotelluric data showing phase rolling out of quadrant (PROQ), in SEG Technical Program Expanded Abstracts 2000 , pp. 344– 346, Society of Exploration Geophysicists. Dentith M.C., Johnson S.P., Evans S., Aitken A.R.A., Joly A., Thiel S., Tyler I.M., 2014. A Magnetotelluric Traverse Across the Eastern Part of the Capricorn Orogen : Geological Survey of Western Australia, Report 135, 49 pp. Egbert G.D., Kelbert A., 2012. Computational recipes for electromagnetic inverse problems. Geophys. J. Int. , 189( 1), 251– 267. https://doi.org/10.1111/j.1365-246X.2011.05347.x Google Scholar CrossRef Search ADS   Groom R.W., Bailey R.C., 1989. Decomposition of magnetotelluric impedance tensors in the presence of local three-dimensional galvanic distortion, J. geophys. Res. , 94( B2), 1913– 1925. https://doi.org/10.1029/JB094iB02p01913 Google Scholar CrossRef Search ADS   Groom R.W., Bailey R.C., 1991. Analytic investigations of the effects of near-surface three-dimensional galvanic scatterers on MT tensor decompositions, Geophysics , 56( 4), 496– 518. https://doi.org/10.1190/1.1443066 Google Scholar CrossRef Search ADS   Heise W., Pous J., 2003. Anomalous phases exceeding 90° in magnetotellurics: anisotropic model studies and a field example, Geophys. J. Int. , 155( 1), 308– 318. https://doi.org/10.1046/j.1365-246X.2003.02050.x Google Scholar CrossRef Search ADS   Ichihara H., Mogi T., 2009. A realistic 3-D resistivity model explaining anomalous large magnetotelluric phases: the L-shaped conductor model, Geophys. J. Int. , 179( 1), 14– 17. https://doi.org/10.1111/j.1365-246X.2009.04310.x Google Scholar CrossRef Search ADS   Ichihara H., Mogi T., Yamaya Y., 2013. Three-dimensional resistivity modelling of a seismogenic area in an oblique subduction zone in the western Kurile arc: constraints from anomalous magnetotelluric phases, Tectonophysics , 603, 114– 122. https://doi.org/10.1016/j.tecto.2013.05.020 Google Scholar CrossRef Search ADS   Jones A.G., 1999. Imaging the continental upper mantle using electromagnetic methods, Lithos , 48( 1-4), 57– 80. https://doi.org/10.1016/S0024-4937(99)00022-5 Google Scholar CrossRef Search ADS   Jones A.G., Groom R.W., Kurtz R.D., 1993. Decomposition and modelling of the BC87 dataset, J. Geomagn. Geoelec. , 45( 9), 1127– 1150. https://doi.org/10.5636/jgg.45.1127 Google Scholar CrossRef Search ADS   Kelbert A., Meqbel N., Egbert G.D., Tandon K., 2014. ModEM: a modular system for inversion of electromagnetic geophysical data, Comput. Geosci. , 66, 40– 53. https://doi.org/10.1016/j.cageo.2014.01.010 Google Scholar CrossRef Search ADS   Krieger L., Peacock J.R., 2014. MTpy: a Python toolbox for magnetotellurics, Comput. Geosci. , 72, 167– 175. https://doi.org/10.1016/j.cageo.2014.07.013 Google Scholar CrossRef Search ADS   Ledo J., Queralt P., Martí A., Jones A.G., 2002. Two-dimensional interpretation of three-dimensional magnetotelluric data: an example of limitations and resolution, Geophys. J. Int. , 150( 1), 127– 139. https://doi.org/10.1046/j.1365-246X.2002.01705.x Google Scholar CrossRef Search ADS   Lezaeta P., Haak V., 2003. Beyond magnetotelluric decomposition: induction, current channeling, and magnetotelluric phases over 90°, J. geophys. Res. , 108( B6), 2305, doi:10.1029/2001JB000990. https://doi.org/10.1029/2001JB000990 Google Scholar CrossRef Search ADS   Lilley F.E.M., Weaver J.T., 2010. Phases greater than 90° in MT data: analysis using dimensionality tools, J. appl. geophys. , 70( 1), 9– 16. https://doi.org/10.1016/j.jappgeo.2009.08.007 Google Scholar CrossRef Search ADS   Martí A., Queralt P., Ledo J., 2009. WALDIM: A code for the dimensionality analysis of magnetotelluric data using the rotational invariants of the magnetotelluric tensor, Comput. Geosci. , 35( 12), 2295– 2303. https://doi.org/10.1016/j.cageo.2009.03.004 Google Scholar CrossRef Search ADS   Martí A., 2014. The role of electrical anisotropy in magnetotelluric responses: from modelling and dimensionality analysis to inversion and interpretation, Surv. Geophys. , 35( 1), 179– 218. https://doi.org/10.1007/s10712-013-9233-3 Google Scholar CrossRef Search ADS   Mccuaig C., Hronsky, J., 2014. The mineral system concept: the key to exploration targeting, pp. 153– 175, BT–Building Exploration Capability for, Society of Economic Geologists, Inc. Pek J., 2009. Effects of electrical anisotropy upon magnetotelluric data: modelling and experiments, in Modern Methods of Electromagnetic Data Measurement, Processing and Interpretation , pp. 110– 135, ed. Spichak V.V., Librokom Publ (in Russian). Pous J., Heise W., Schnegg P., Muñoz G., Martí J., Soriano C., 2002. Magnetotelluric study of the Las Cañadas caldera (Tenerife, Canary Islands): structural and hydrogeological implications, Earth planet. Sci. Lett. , 204( 1–2), 249– 263. https://doi.org/10.1016/S0012-821X(02)00956-1 Google Scholar CrossRef Search ADS   Selway K., Sheppard S., Thorne A., Johnson S., Groenewald P., 2009. Identifying the lithospheric structure of a Precambrian Orogen using magnetotellurics: the Capricorn Orogen, Western Australia, Precambrian Res. , 168( 3–4), 185– 196. https://doi.org/10.1016/j.precamres.2008.09.010 Google Scholar CrossRef Search ADS   Selway K., Thiel S., Key K., 2012. A simple 2-D explanation for negative phases in TE magnetotelluric data, Geophys. J. Int. , 188( 3), 945– 958. https://doi.org/10.1111/j.1365-246X.2011.05312.x Google Scholar CrossRef Search ADS   Sheppard S., Fletcher I.R., Rasmussen B., Zi J., Muhling J.R., Occhipinti S.A., Wingate M.T.D., Johnson S.P., 2016. A new Paleoproterozoic tectonic history of the eastern Capricorn Orogen, Western Australia, revealed by U–Pb zircon dating of micro-tuffs, Precambrian Res. , 286, 1– 19. https://doi.org/10.1016/j.precamres.2016.09.026 Google Scholar CrossRef Search ADS   Simpson F., Bahr K., 2005. Practical Magnetotellurics , Cambridge Univ. Press. Google Scholar CrossRef Search ADS   Thiel S., Heinson G., Gray D.R., Gregory R.T., 2009. Ophiolite emplacement in NE Oman: constraints from magnetotelluric sounding, Geophys. J. Int. , 176( 3), 753– 766. https://doi.org/10.1111/j.1365-246X.2008.04053.x Google Scholar CrossRef Search ADS   Weaver J.T., Agarwal A.K., Lilley F.E.M., 2000. Characterization of the magnetotelluric tensor in terms of its invariants, Geophys. J. Int. , 141( 2), 321– 336. https://doi.org/10.1046/j.1365-246x.2000.00089.x Google Scholar CrossRef Search ADS   Weckmann U., Ritter O., Haak V., 2003. A magnetotelluric study of the Damara Belt in Namibia, Phys. Earth planet. Inter. , 138( 2), 91– 112. https://doi.org/10.1016/S0031-9201(03)00079-7 Google Scholar CrossRef Search ADS   SUPPORTING INFORMATION Supplementary data are available at GJI online. Figure S1. Example of sites with anomalous phases recorded in the study area (dots and triangles) and its comparison with the ongoing 3-D inversion models (lines) for both, off-diagonal (left) and diagonal (right) components. Figure S2. Phase tensor ellipses (Phi.min) and induction arrows (black: real; red: imaginary) for each BBMT site at six different periods. Figure S2. 3-D synthetic models used to try and explain the observed POQ and calculated phases curves for some sites with anomalous YX phases. Dots and triangles: observed phases; Lines: Synthetic phases. (a) General background model (Model S) for 0–8.4 km depth; at greater depths the model is a homogeneous half-space of 10000 Ωm. (b) Set of models with a regional E–W trending conductive body at depth (8.4–9.8 km; resistivity: 2 Ωm). (c) Model S-21, same structure than Model S-7 with a local conductive body of 10 Ωm in the centre of the Marymia Inlier from 0.5 to 8.4 km depth. (d) Set of models performed modifying the thickness of the basins to the east. The southernmost basin is closed at depths greater than 1.8 km. Model S-3 is equivalent to Model S, without regional E–W conductive body; Model S3.7 is equivalent to Model S-7, with the same regional E–W conductive body. (e) Tests performed to study the effect of the resistivity distribution in the Palaeoproterozoic basin in Model S-7, since resistivities larger than 50 Ωm are expected below 2 km depth. Layered basins with resistivity values increasing with depth provide similar results in terms of anomalous phases at long periods. Please note: Oxford University Press is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the paper. © The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Geophysical Journal International Oxford University Press

Magnetotelluric data from the Southeastern Capricorn Orogen, Western Australia: an example of widespread out-of-quadrant phase responses associated with strong 3-D resistivity contrasts

Loading next page...
 
/lp/ou_press/magnetotelluric-data-from-the-southeastern-capricorn-orogen-western-eRvd06LGah
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society.
ISSN
0956-540X
eISSN
1365-246X
D.O.I.
10.1093/gji/ggx459
Publisher site
See Article on Publisher Site

Abstract

Summary Out-of-quadrant impedance phases (POQ) have been observed in several magnetotelluric (MT) data sets around the world, hampering the modelling and interpretation of the data. These anomalous responses are usually observed in small groups of sites and have been variously interpreted as due to electrical anisotropy, galvanic distortion, 2-D structures with large resistivity contrasts, or 3-D conductive bodies, such as L-shaped conductors. We present here what is possibly a unique land-based MT data set characterized by an exceptionally large number of sites with POQ. The MT data were collected in the southeastern margin of the Capricorn Orogen (Western Australia). This is an area with a complex geological history, which is evident in the MT responses. Given the characteristics of the study area, which includes very high resistivity Archean terrains surrounded by low resistivity Palaeoproterozoic basins, strong resistivity contrasts involving complex 3-D structures are a possible cause of this unusual behaviour. This is confirmed by 3-D forward modelling using models based on the general geology of the study area, as well as by the 3-D inversion results. A comparison of the real and synthetic data, derived from 3-D forward modelling, using the same parameters suggests that the proposed scenario is a realistic explanation for most of the anomalous phases observed in this data set. Synthetic data are not affected by galvanic distortion or electrical anisotropy, so the good match observed between the real and synthetic MT responses is likely due to 3-D inductive effects. Previous explanations for POQ responses, such as blocks with electrical anisotropy and 3-D L-shaped conductors, are not required and are considered to be less likely causes of the POQ responses on geological grounds. Australia, Magnetotellurics, Cratons 1 INTRODUCTION A large multi-disciplinary geoscience project ‘Distal Footprints of Giant Ore Systems’ is being undertaken in the Capricorn Orogen, in the central part of Western Australia. The project involves large scale geophysical data acquisition including passive seismic recordings, airborne electromagnetics and a major magnetotelluric (MT) study (Aitken et al. 2015). A key objective of the deep geophysical component of the project is to map the boundaries of major Archean basement blocks beneath the cover of younger sedimentary rocks, because these are considered an important indicator of regional prospectivity for several important kinds of mineral deposits (McCuaig & Hronsky 2014). An unusual characteristic of the MT data set from the Capricorn Orogen is the large number of sites with phases out-of-quadrant (POQ) at long periods. The phases for the off-diagonal components of the impedance tensor will normally lie in the first (0° < phs. <90°) and third quadrant (–90° < phs. < –180°) for the XY and YX components, respectively. The authors are not aware of such a large number of sites with POQ (56 sites of a total of 83 measured across an area of about 40 000 km2) observed elsewhere in the world, possibly making this a exceptional MT data set. Although uncommon, POQ MT responses have been observed in other data sets collected elsewhere in the world and several studies have attempted to explain the origin of this phenomenon. POQ has been explained by the presence of electrical anisotropy (e.g. Heise & Pous 2003; Pek 2009), galvanic distortion (e.g. Chouteau & Tournerie 2000; Lilley & Weaver 2010), 2-D structures with large resistivity contrasts (e.g. Selway et al. 2012), or 3-D conductive bodies (e.g. Lezaeta & Haak 2003; Ichihara & Mogi 2009; Thiel et al. 2009; Ichihara et al. 2013). Anomalous phases due to electrical anisotropy appear mainly in sites located above an anisotropic block and near its borders (e.g. Heise & Pous 2003). 3-D models proposed to explain POQ generally include conductive structures with contact angles of 90°, such as L-shaped bodies (Ichihara & Mogi 2009) or a hollow square (e.g. Pous et al. 2002; Weckmann et al. 2003). In the L-shaped bodies case, the locations of the POQ responses are restricted to the corner where the two conductors that make up the L-shape are connected, whilst in the hollow square case POQ occur in recordings directly above the conductive body, but not inside or outside of the body. Thus, there are clear restrictions on the locations of the sites with anomalous phases relative to the conductive body that is causing the POQ. Such local-scale explanations are very unlikely to be applicable to the Capricorn Orogen data set because of the large number of sites involved and the fact that these occur over an area of around 200 km × 200 km. Seeking an alternative explanation, we have attempted to explain this phenomenon using 3-D forward modelling of data from the southeastern part of the Carpicorn Orogen (Fig. 1). Figure 1. View largeDownload slide (a) Location of the study area. (b) Location of the sites with anomalous YX phases (red triangles) independent of observation angle. (c) Location of the sites with anomalous YX phases (red triangles) for a specific observation angle, rotation of the impedance tensor equal to zero (see text for more information). Figure 1. View largeDownload slide (a) Location of the study area. (b) Location of the sites with anomalous YX phases (red triangles) independent of observation angle. (c) Location of the sites with anomalous YX phases (red triangles) for a specific observation angle, rotation of the impedance tensor equal to zero (see text for more information). 2 GEOLOGICAL SETTING The Proterozoic Capricorn Orogen is a zone of deformed and metamorphosed igneous and sedimentary rocks located in between the Archaean Pilbara and Yilgarn Cratons (Fig. 1a). The Orogen comprises the deformed margins of the two Cratons, the Archaean–Palaeoproterozoic Gascoyne Province and a series of Palaeoproterozoic–Mesoproterozoic sedimentary basins, for example the Collier, Yerrida and Bryah Basins (Fig. 1). There is an inlier of Yilgarn Craton granitoid-greenstone rocks, the Marymia Inlier, immediately to the north of the Yilgarn Craton. The Capricorn Orogen has a very long and complex tectono-magmatic history related to the collision of several continental basement terranes and also subsequent intraplate deformation events, as described by Sheppard et al. (2016). 3 MT DATA SET A total of 83 broad-band magnetotelluric stations (BBMT) have been recorded to date in the southeast part of the Capricorn Orogen (mostly within the Bryah-Padbury and Yerrida Basins) and adjacent Yilgarn Craton (Fig. 1). This data set will be referred herein as the SECO-data set. The field area is remote and difficult to access. Logistical constraints required the data to be collected as traverses along roads and tracks. The MT data acquisition was undertaken in three stages. The easternmost sites were recorded in May 2011 (Dentith et al. 2014), while the westernmost sites were acquired in October 2013. Twenty new BBMT sites were collected in October 2015, mostly in the Collier Basin around the Marymia Inlier. Data were acquired using Metronix ADU07e and Phoenix MTU-5A recorders with induction coils, and non-polarized electrodes (Pb/PbCl2) with a typical dipole length of 100 m. The x-axis was oriented in the magnetic N–S direction, with the positive direction pointing to the north, and the y-axis in the E–W direction, positive to the east. The data were recorded at each location for an average of 48 hr, and the typical spacing between stations is 10 km. Archean rocks in the upper crust are expected to have high resistivities (e.g. Jone 1999; Selway et al. 2009), whilst the Proterozoic basins are expected to have lower resistivity values, due to the presence of shales and iron-rich formations. These strong resistivity contrasts and the complicated outcrop distribution of basins and basement, resulting from the area's complex geological history, are reflected in the complexity of the MT data. 3.1 Phases out of quadrant An important characteristic of the SECO-data set is the large number of sites with phases that are out of quadrant (above 90°) at long periods (Fig. S1). This behaviour is typically observed at periods longer than 10 s and across the whole study area, and there is no obvious relationship between the location of the sites with anomalous phases and the outcropping geology (Figs 1 and S1). This effect is strongly dependent on the rotation angle of the impedance tensor (Z). Z is a complex tensor which relates the orthogonal electric and magnetic fields recorded at each MT site. Thus, it contains information about dimensionality and direction (Simpson & Bahr 2005). A mathematical rotation of the impedance tensor from its measuring axis will provide information about the electrical dimensionality of the data, but in our particular case, it also highlights a strong dependence of the occurrence of anomalous phases on this rotation angle. Data were recorded with the x-axis in the magnetic N–S direction. This corresponds to a rotation angle of 0°, whilst a rotation angle of 90° will lead to the x-axis being parallel to the E–W direction. This rotation modifies all components of Z and some of the components can become zero (e.g. 2-D structures where one of the axes is aligned along the electrical strike, in which case the off-diagonal components of Z are zero). However, regardless of dimensionality (1-D, 2-D, 3-D structures) and rotation angle, phases are not expected to exceed the ordinal quadrant limits. Fig. 1(b) shows the location of the sites with YX-POQ (in red) independent of rotation angle, that is those sites for which the YX phases are out of quadrant for any rotation angle between 0° and 180°. Fig. 1(c) shows the location of the sites with YX-POQ for a rotation angle for the impedance tensor equal to zero. Note the large number of distorted sites located in, and around, the Marymia Inlier. We concentrate on stations from this area (grey box) in our analysis of the POQ. Fig. 2 shows the dependence of the phase response on the impedance rotation angle using the YX phases for site MRY022 (yellow circle in Fig. 2b). The purple region in Fig. 2(a) shows the periods in which the phases are outside the third quadrant (Y-axis) for the different rotation angles (X-axis). The black and red vertical lines represent the limits of this out of quadrant range of angles, for the specific period of 10 s (0.1 Hz, white dashed line). This is in general the lowest period where POQ occurs across the study area. A clockwise rotation from the black line keeps the phases within the ordinal quadrant. The same effect is observed with an anticlockwise rotation from the red line. Figure 2. View largeDownload slide (a) YX phases for site MRY022 according to the observation angle. White dashed line indicates a period of 10 s. Black and red lines correspond to the limits of the area where the phases are within the third quadrant, for the period chosen. (b) Range of angles for which the YX phases are within the ordinal quadrant (in grey), for a period of 10 s. Yellow circle: Site MRY022. Surface geology is in the background. (c) Orientation of the electric field (blue lines) for site MRY022 when the phases are out of quadrant, for a period of 10 s. Red and black lines: range of angles for which the YX phases are out the third quadrant. Black dashed line: Trend of the Marymia Inlier limits. Figure 2. View largeDownload slide (a) YX phases for site MRY022 according to the observation angle. White dashed line indicates a period of 10 s. Black and red lines correspond to the limits of the area where the phases are within the third quadrant, for the period chosen. (b) Range of angles for which the YX phases are within the ordinal quadrant (in grey), for a period of 10 s. Yellow circle: Site MRY022. Surface geology is in the background. (c) Orientation of the electric field (blue lines) for site MRY022 when the phases are out of quadrant, for a period of 10 s. Red and black lines: range of angles for which the YX phases are out the third quadrant. Black dashed line: Trend of the Marymia Inlier limits. Fig. 2(b) shows the angles for which the YX phases are within and out of the third quadrant, for each site for a period of 10 s. Most of the sites with anomalous phases at this period are located in the northern part of the study area. 4 DATA ANALYSIS MT tensor analysis was performed on the SECO-data set using several different techniques, in order to understand the dimensionality of the regional electrical variations, as well as try to identify galvanic and inductive causes of the anomalous phases, since this anomalous behaviour has been associated with both causes (see Section 1). The analysis reveals clear 3-D geoelectrical behaviour and extreme complexity for most of the sites recorded in this area. Given this, several 3-D electrically isotropic synthetic models were created to try and reproduce the anomalous phases, since 3-D complex structures involving strong resistivity contrasts are considered as a potential cause of the POQ. Comparison of the observed and synthetic MT responses from those sites with anomalous phases helps to understand the mechanisms related to this unusual behaviour since, for example, the synthetic data are not affected by galvanic distortion. In this section, we show the results of the data analysis for both observed and synthetic responses, while their comparison is addressed in the discussion. 4.1 Observed data 4.1.1 Phase tensors Fig. 3 shows the phase tensor (Caldwell et al. 2004), computed using the MTpy package (Krieger & Peacock 2014), for each station for each of six different periods from 0.01 to 600 s. The phase tensor is one of the most common techniques to analyse MT data, since it is not affected by galvanic distortion and does not make any assumptions about the dimensionality of the regional electrical variations (Caldwell et al. 2004). Its graphical representation is as an ellipse, where both axes are equal (circle) in a 1-D medium, whilst in a 2-D case the axes are parallel or perpendicular to the strike of the structure. For 3-D structure the phase tensor is non-symmetric and the skew is the angle which gives the rotation of the major axis away from an identically shaped ellipse (Chave & Jones 2012). Skew values larger than ±5 are usually considered to be an indication of a 3-D resistivity distribution. Figure 3. View largeDownload slide Phase tensor ellipses for each BBMT site at six different periods. Figure 3. View largeDownload slide Phase tensor ellipses for each BBMT site at six different periods. The results of this analysis show the obvious three-dimensionality of the data, with skew values reaching greater than ±10 at periods greater than 1 s, while at shorter periods, skew values are usually lower than ±5. One of the most intriguing aspects revealed by Fig. 3 is the behaviour of the data at sites located in the vicinity of the Marymia Inlier (grey box in Fig. 1c). At 1s, there is a group of five sites with a skew lower than –10 and the major axis of the phase tensor aligned in a NW–SE direction. At 10s, some neighbouring sites are included in this group, ten in total, but in this case the major axis of the phase tensor has a slight clockwise rotation relative to the 1 s data. At 100 s the group comprises thirteen sites, all of them with similar high skew values. The major axis of the phase tensor shows a further clockwise rotation, being now in a NNE–SSW direction. Finally, at 600 s this group of sites with similar behaviour is no longer apparent, even though elsewhere in the area the responses at 100 and 600 s are quite similar. 4.1.2 WAL invariants An analysis of the WAL invariants (Weaver et al. 2000) was performed using the WALDIM code (Martí et al. 2009). This method also provides information about the dimensionality and the orientation of the electrical structure. As observed using the phase tensors, many of the data behave three-dimensionally at periods longer than 1 s (not shown here). There is a set of eight rotational invariants known as WAL invariants, seven independent (I1 to I7) and one dependent (Q). Invariants I3 to I7 and Q can be used to determine the type of dimensionality and identify galvanic distortion (Martí et al. 2009). Invariants I7 and Q provide information about the dimensionality, since I7 is non-zero only for 3-D cases and Q is zero only for 1-D. Invariants I5, I6 and I7 are related to galvanic distortion, while I3 and I4 have information about the degree of 2-D anisotropy (Chave & Jones 2012). Fig. 4 shows the values of the invariants I3 to I7 for two sites from the SECO-data set. Fig. 4(a) shows the WAL invariant of one site with POQ (see Fig. 1c for location), while Fig. 4(b) shows the same invariants for one of the sites with phases always within their respective quadrants regardless of the rotation angle (see Fig. 1c for location). Both sites show non-zero values of the invariants I3, I4 and I7 for long periods, indicating the 3-dimensionality of the data. Figure 4. View largeDownload slide WAL invariants I3, I4, I5, I6 and I7. (a) Example of a site with anomalous YX phases. (b) Example of a site with YX phases in the ordinal quadrant (see Fig. 1c for location, open circles). Figure 4. View largeDownload slide WAL invariants I3, I4, I5, I6 and I7. (a) Example of a site with anomalous YX phases. (b) Example of a site with YX phases in the ordinal quadrant (see Fig. 1c for location, open circles). The main difference between the sites with and without POQ is in the invariants I5 and I6. Sites with phases within the ordinal quadrant (Fig. 4b) have relatively low values for the invariant I6 for the whole period range, while sites with POQ (Fig. 4a) show higher values at long periods. Meanwhile, invariant I5 is very small at high frequencies for both types of site, as well as at long periods for those sites with phases within the ordinal quadrant. For those sites with POQ the value of this invariant at long periods is clearly non-zero. 4.1.3 Groom and Bailey decomposition The Groom and Bailey decomposition technique is used to separate local and regional effects in the MT responses, and resolve the geoelectrical strike. There are four independent parameters in the distortion tensor, including three tensor suboperators: twist, shear and anisotropy, and one scalar: site gain, which is the only indeterminable parameter. Unlike the phase tensor and WAL invariants, which make no assumptions about the dimensionality, this technique assumes that the Earth has a 2-D regional conductive structure. Therefore, large 3-D inductive effects are not accounted for directly in the method. Nonetheless, this analysis can provide some information about the presence of 3-D induction, since the twist and shear parameters will show frequency dependence under these circumstances (Simpson & Bahr 2005). Fig. 5 shows the unconstrained Groom & Bailey (1989, 1991) galvanic distortion parameters twist and shear for the sites with POQ (Fig. 5a), and without POQ (Fig. 5b). There are clear differences between these two groups of data. Sites with POQ show very high values of shear, rising quickly to its physical limit, ±45°. Such values are normally regarded as severe current channelling effects, for example Jones et al. (1993). For this group of sites, the twist shows an approximate frequency-independent behaviour at short periods with a rapid increase in its value at periods greater than 1 s. In contrast, sites with phases within the ordinal quadrant show less extreme behaviour (Fig. 5b). In general, the shear is lower than 30° and never reaches the limit of ±45°. The twist is scattered at periods longer than 10 s, but it does not show the steep slopes observed in the sites with POQ (Fig. 5a). Figure 5. View largeDownload slide Groom and Bailey parameters for the sites located in the Marymia Inlier area (see Fig. 1c for location, grey box). (a) Sites with YX anomalous phases. (b) Sites with YX phases within the third quadrant. Figure 5. View largeDownload slide Groom and Bailey parameters for the sites located in the Marymia Inlier area (see Fig. 1c for location, grey box). (a) Sites with YX anomalous phases. (b) Sites with YX phases within the third quadrant. 4.2 Forward modelling Given the unusual number of sites with POQ that are located across a large area, the distinctive 3-D geoelectrical behaviour demonstrated by the various data analysis methods described above, and the expected strong resistivity contrasts between the Archean units and the Palaeoproterozoic basins, we consider 3-D variations involving strong resistivity contrasts as a possible cause of the anomalous phases. In order to analyse this possibility, several 3-D electrically isotropic synthetic models were created to try and reproduce the observed POQ. Due to the variety and complexity in the responses observed in the SECO-data set, we focus the analysis on those data from the vicinity of the Marymia Inlier (grey box in Fig. 1c). Forward modelling was carried out using the ModEM code (Egbert & Kelbert 2012; Kelbert et al. 2014). The responses have been calculated for the full impedance tensor for 18 periods, in 61 sites located according to the actual MT sites. All the models have the same basic structure, which consists of Archean units with resistivities of 10 000 Ωm and basins with resistivities of 50 Ωm. The basins are assumed to be 8.4 km deep, with Archean basement beneath, also with a resistivity of 10000 Ωm (Fig. 6). This overall structure is derived from the surface geology, while the resistivity values are assigned taking into account the actual MT data. Variants on this simplest model (Model S) were created mainly by adding conductive bodies at different depths and in different areas, as well as reducing the thickness of the youngest Palaeoproterozoic basin (the Yerrida basin). Fig. S2 shows some of the models tested, as well as the responses predicted in comparison to the observed MT data. Figure 6. View largeDownload slide Example of two 3-D synthetic models used to try and explain the observed POQ (more models are shown in Fig. S2). Model S: General background model for 0–8.4 km depth; at greater depths the model is a homogeneous half-space of 10 000 Ωm. Model S-7: Same shallow structure than Model S, but with an E–W trending conductive body (2 Ωm) eastward the Marymia Inlier at depth (8.4– 9.8 km). Red triangles, sites with XY anomalous phases; Yellow triangles, sites with YX anomalous phases. Figure 6. View largeDownload slide Example of two 3-D synthetic models used to try and explain the observed POQ (more models are shown in Fig. S2). Model S: General background model for 0–8.4 km depth; at greater depths the model is a homogeneous half-space of 10 000 Ωm. Model S-7: Same shallow structure than Model S, but with an E–W trending conductive body (2 Ωm) eastward the Marymia Inlier at depth (8.4– 9.8 km). Red triangles, sites with XY anomalous phases; Yellow triangles, sites with YX anomalous phases. Several models, including the simplest one (Model S, Fig. 6), lead to phases being out of quadrant; but only those models including a conductive body eastward of the Marymia Inlier induce the YX phases (the main component affected in the real data set) to leave the third quadrant (e.g. Model S-7, Fig. 6). This demonstrates that strong 3-D resistivity contrasts are possibly associated with the POQ observed in complex geological contexts, such as our study area. 4.2.1 Synthetic MT responses Fig. 7 shows the phase tensor, WAL invariants and Groom and Bailey parameters calculated from the responses of the synthetic models S and S-7 (Fig. 6). This corresponds to those sites where POQ has been reproduced, such as site MRY22 (Figs 7a and c). As observed in the SECO-data set, the phase response is also strongly dependent on the observation angle (compare Fig. 7a with Fig. 2a). Figure 7. View largeDownload slide Synthetic model responses. (a) YX phases for the site MRY022 (response from synthetic Model S-7) according to the observation angle. (b) Phase tensor ellipses at 1s period for the sites located on and around the Marymia Inlier (data from synthetic Model S). (c) WAL invariants I3, I4, I5, I6 and I7 for the site MRY022 (response from synthetic Model S-7). (d) Groom and Bailey parameters for the three sites with YX anomalous phases obtained in the Model S-7 (see Fig. 6 for location). Figure 7. View largeDownload slide Synthetic model responses. (a) YX phases for the site MRY022 (response from synthetic Model S-7) according to the observation angle. (b) Phase tensor ellipses at 1s period for the sites located on and around the Marymia Inlier (data from synthetic Model S). (c) WAL invariants I3, I4, I5, I6 and I7 for the site MRY022 (response from synthetic Model S-7). (d) Groom and Bailey parameters for the three sites with YX anomalous phases obtained in the Model S-7 (see Fig. 6 for location). Fig. 7(b) shows the phase tensor at 1 s for Model S in the vicinity of the Marymia Inlier. Here, the SECO-data set shows a clear orientation of the major axis of the phase tensor, perpendicular to the trend of the inlier (Fig. 3). At 1s the major axis of the phase tensor in those sites located on the inlier has the same orientation as observed in the real data, although we have not been able to reproduce the same clockwise rotation with increasing period observed in the SECO-data set. The SECO-data set shows a clear difference in the WAL invariants between the sites with phases in the ordinal quadrant and those sites with POQ (Fig. 4). The responses from the synthetic models show a similar behaviour for this last group of sites: non-zero values for all the invariants represented (I3 to I7) at long periods (compare Fig. 7c with Fig. 4a). Note the good correlation, in both values and trend, of all the invariants at long periods. Fig. 7(d) shows the unconstrained twist and shear angles of the Groom and Bailey decomposition for Model S-7 for those sites with POQ (see Fig. 6 for location). As observed in the SECO-data set (Fig. 5a) the shear rises quickly to 45°, while the twist shows an approximate frequency-independent behaviour at short periods with a rapid increase in its value at periods longer that 1s. 5 DISCUSSION The most remarkable characteristic of the SECO-data set is the large number of sites with anomalous phases. This unusual behaviour has been related to electrical anisotropy (e.g. Heise & Pous 2003; Pek 2009), galvanic distortion (e.g. Chouteau & Tournerie 2000; Lilley & Weaver 2010), 2-D structures with large resistivity contrasts (e.g. Selway et al. 2012), or 3-D conductive bodies (e.g. Lezaeta & Haak 2003; Ichihara & Mogi 2009; Thiel et al. 2009; Ichihara et al. 2013). Among these possibilities, we consider the 3-D structures with strong resistivity contrasts as the most likely cause of the POQ. This assumption is based on the geological characteristics of the study area, with Archean units surrounded by Palaeoproterozoic basins (e.g. Marymia Inlier, see Fig. 1) resulting in complex 3-D structures with potentially strong resistivity contrasts, as well as the results obtained from the data analysis, which reveals the area to be 3-D from the geoelectrical point of view. Other significant characteristic of the SECO-data set is that the anomalous phases show a strong sensitivity to the rotation angle of the impedance tensor (Figs 1 and 2). There is a range of angles for which the phases remain within their respective quadrants. Such behaviour has been observed in other studies, for example Jones et al. (1993) and Heise & Pous (2003), corresponding to those angles with the direction of the structural strike, in the case of 2-D isotropic models, or with the anisotropy strike, in the case of 2-D anisotropic models. A more detailed analysis of Fig. 2(b) in the Marymia Inlier area, highlights a correlation between the range of angles for which the phases are within the ordinal quadrant and the average trend of the inlier itself. Additionally, Fig. 2(c) shows the orientation of the electric field for the rotation angles of the impedance tensor for which the YX phases are out of the third quadrant. There is also a clear relationship between the trend of the inlier and the electric field orientation for those observation angles. Thus, YX phases are running out of quadrant when the electric field is approximately parallel to the inlier. All this can be interpreted as evidence that the strong resistivity contrasts expected along the contacts between the Archean basement and Palaeoproterozoic basins play an important role in the distortion that causes the phases to run out of quadrant. A similar behaviour has been observed in 2-D structures with a strong resistivity contrast between the basement and the cover (e.g. Selway et al. 2012). Thus, similar mechanisms could be involved in the responses observed here, although a 3-D approach is required due to both the characteristics of the MT data observed and the geological setting. This idea of strong 3-D resistivity contrast as main cause of the POQ is supported by the good agreement observed between the MT responses of the observed and synthetic data. Firstly, the orientation of the phase tensors at periods greater than 1 s (Fig. 3) could be largely due to the characteristics of the contact between the Marymia Inlier and the Palaeoproterozoic basins, as demonstrated by the synthetic model responses (Fig. 7b). Secondly, extremely high values of shear (Figs 4 and 7d) are typically indicative of severe current channelling effects at these periods (e.g. Jones et al. 1993). As the ± 45° limit is approached, the electrical field is basically propagating in one direction (Chave & Jones 2012). Both the real and synthetic data sets are characterised by a rapid rise in the shear values, reaching ±45° at long periods. The synthetic data are not affected by galvanic distortion, thus the strong period dependence of the shear and twist at periods between 1 and 100 s (Fig. 7D) indicates there are inductive effects associated with the E–W conductive body included in Model S-7 (e.g. Ledo et al. 2002). The similarity in the decomposition parameters between the synthetic data and the SECO-data set is an indication that both could be distorted by similar mechanisms. Therefore, 3-D inductive effects would be directly related to these anomalous phases, thus discarding the galvanic distortion as a potential cause. Finally, the dimensionality criteria according to the WAL invariants values, suggests a three-dimensionality affected, or not, by galvanic distortion for both real and synthetic data sets. Lilley & Weaver (2010) used the rotational invariants to analyse the anomalous phases in a case of galvanic distortion of a 2-D structure. Evidence for the anomalous phases observed in the SECO-data set being the result of galvanic distortion cannot be inferred from the findings presented here. WAL invariants from synthetic data with POQ (Fig. 7c) show a comparable distribution to that observed in the real data (Fig. 4a). Given this similarity, both data sets could be distorted by similar mechanisms, which do not include galvanic distortion or electrical anisotropy, both absent in the synthetic data. In summary, all the parameters analysed point to inductive effects involving 3-D structures as the primary cause of most of the POQ recorded in the SECO-data set. This effect arises due to the presence of various 3-D conductive bodies in combination with the strong resistivity contrast expected at the interface between resistive Archean rocks and the less resistive rocks of the Palaeoproterozoic basins. 5.1 3-D inversion model Taking into account the results obtained from the data analysis, 3-D inversion techniques were considered appropriate tools to perform the modelling of the SECO-data set. Furthermore, the interpretation of an inductive cause for the POQ phenomenon would be supported if the inversion models can achieve a good fit to the observed data with anomalous phases. 3-D inversion models were computed with the ModEM code (Egbert & Kelbert 2012: Kelbert et al. 2014). No rotation of the impedance tensor was applied to the data for the inversion, which used the four components of the impedance tensor. Initially, only 35 sites with phases in the ordinal quadrants for the observation angle were included in the inversion process, for 12 periods and covering the whole period range (stage 1). Subsequently, sites with anomalous phases were added to the inversion (58 sites in total), but only for those periods unaffected by the anomalous phases (T < 1 s) (stage 2). Finally, those periods for which the phases run out-of-quadrant were included in the inversion (stage 3). Fig. 8 shows map views of the 3-D inversion model at different depth for the three stages of inversion mentioned above. Comparison of the models allows recognition of aspects of the model required to fit the anomalous phases. The main differences are observed at depth, where the resistivity features labelled HR and LR appear better defined and with more lateral continuity. This is consistent with the synthetic models performed to reproduce POQ, where the anomalous phases are mainly due to the strong resistivity contrast expected between the Archean units and the Palaeoproterozoic basins (Figs 6 and S2 in SM). New sites with POQ appear by adding conductive bodies in the basins or modifying the limits of the Archean (Fig. S2). Thus, both inversion and synthetic models are coherent regarding the main characteristics (e.g. resistivity value, lateral extension) of the main structures related to the POQ. Figure 8. View largeDownload slide Map views of the 3-D inversion models at different depths. Stage 1: Inversion performed with 35 sites for the whole period range; no POQ are included in the inversion process. Stage 2: Inversion performed with 58 sites for periods shorter than 1 s; no POQ are included in the inversion process. Stage 3: Inversion performed with 58 sites for the whole period range; POQ are included in the inversion process. Figure 8. View largeDownload slide Map views of the 3-D inversion models at different depths. Stage 1: Inversion performed with 35 sites for the whole period range; no POQ are included in the inversion process. Stage 2: Inversion performed with 58 sites for periods shorter than 1 s; no POQ are included in the inversion process. Stage 3: Inversion performed with 58 sites for the whole period range; POQ are included in the inversion process. A good data fit has been achieved in the inversion performed with 58 sites for the whole period range, even for those sites with anomalous phases. Fig. S1 shows the comparison between the observed and predicted responses (stage 3 model) for some sites with POQ, for the four components of the impedance tensor. Thus, the results obtained from the inversion models support the idea of strong 3-D resistivity contrasts as a main cause of the anomalous phases. 5.2 Electrical anisotropy Given the geological complexity of the study area we cannot ignore the possibility of there being electrical anisotropy affecting the SECO-data set. To explain the POQ as a result of electrical anisotropy an outcropping anisotropic block with anisotropic azimuth different to the structural azimuth is required (e.g. Heise & Pous 2003; Pek 2009). The model proposed by Heise & Pous (2003) consists of an anisotropic block underlain by an anisotropic layer with anisotropy strikes perpendicular to each other, whilst in the model proposed by Pek (2009), only the anisotropic block is needed (Marti 2014). These models require an outcropping anisotropic block and all the sites with anomalous phases must be located above it. Outside the block, the distortion effects do not cause the phases to be out of quadrant. In view of these requirements, the large number and location of the sites with POQ (Fig. 1), and the good agreement between the responses of the SECO-data set and the synthetic models; we consider it unlikely that electrical anisotropy is the primary cause of the POQ. 6 CONCLUSIONS We present here what is possibly a unique land-based MT data set with a remarkable number of recording sites with phases out of quadrant. This effect is observed mainly in the YX component and extends across the whole study area, an area of around 40 000 km2. Here we summarize the main findings: Dependence on the observation angle: One of the most striking characteristics of this data set is the dependence observed between the occurrence of POQ and the rotation angle of the impedance tensor. The main direction of the structures that generate strong resistivity contrasts seems to be controlling the range of angles for which the phases are within the ordinal quadrant. Therefore, the phases come into their respective quadrants as the observation angle approaches this direction. Thus, the emergence of anomalous phases is strongly dependent on the relationship of both, the observation angle (Z rotation angle) and the trend of the structure that generates the strong resistivity contrast. 3-D inductive effects as a main cause of POQ. This is based on: The 3-D forward models and the good agreement between the MT responses from both real and synthetic data, show that the observed POQ are most likely caused by strong resistivity contrasts with complex 3-D geometries. 3-D inversions achieve a good data fit, even for the anomalous phases. This is a key result in the inversion of this particular data set, which can be interpreted as indicative of the inductive nature of the POQ phenomenon in this area. Geometry of the 3-D structures: It is important to note that 3-D structures with angles of exactly 90 degrees are absent in the synthetic models here presented. Previous work reproducing POQ with 3-D models include conductive bodies with 90 degree angles, such as L-Shaped bodies or hollow squares (e.g. Pous et al. 2002; Weckmann et al. 2003; Ichihara & Mogi 2009). This opens up numerous possibilities to explain anomalous phases as caused by complex 3-D structures. Electrical anisotropy: The presence of electrical anisotropy remains possible but we do not consider that this is the primary cause of such a large number of sites with anomalous phases. If some sites with POQ are recorded in areas with anisotropy, the anisotropic azimuth would probably be included in the range of angles for which the phases are in the ordinal quadrant (Fig. 2b). POQ appear when the observation angle is different from both the direction of the structures that generate strong resistivity contrasts (as shown in this study) and the anisotropic strike (e.g. Pek 2009; Heise & Pous 2003). This needs to be addressed in future work, in order to better understand the MT responses observed in this data set, since in some sites the anomalous phases are likely due to a combination of effects. Acknowledgements This work was supported by resources provided by the Pawsey Supercomputing Centre with funding from the Australia Government and Government of Western Australia. Authors thank Gary Egbert and Anna Kelbert for providing the ModEM code, as well as Naser Meqbel for providing the 3D-Grid visualizer. Moombarriga Geoscience collected the data described here in some cases with the assistance of personnel from the Geological Survey of Western Australia and the Centre for Exploration Targeting. The assistance of Ray Addenbrooke deserves particular acknowledgement. We also thank the various stations owners and the Plutonic Gold Mine for access to the field area. The editor, Ute Weckmann, and the reviewers, Kate Selway and Stephan Thiel, are thanked for the useful comments, which helped to improve the manuscript. REFERENCES Aitken A.et al.  , 2015. A major geophysical experiment in the Capricorn Orogeny, Western Australia, in Proceedings of the ASEG Extended Abstracts 2015: 24th International Geophysical Conference and Exhibition , pp. 1– 5, Australian Society of Exploration Geophysicists. Google Scholar CrossRef Search ADS   Caldwell T.G., Bibby H.M., Brown C., 2004. The magnetotelluric phase tensor, Geophys. J. Int. , 158( 2), 457– 469. https://doi.org/10.1111/j.1365-246X.2004.02281.x Google Scholar CrossRef Search ADS   Chave A.D., Jones A.G., 2012. The Magnetotelluric Method , pp. 1–584, eds Chave A.D., Jones A.G., Cambridge University Press. Google Scholar CrossRef Search ADS   Chouteau M., Tournerie B., 2000. Analysis of magnetotelluric data showing phase rolling out of quadrant (PROQ), in SEG Technical Program Expanded Abstracts 2000 , pp. 344– 346, Society of Exploration Geophysicists. Dentith M.C., Johnson S.P., Evans S., Aitken A.R.A., Joly A., Thiel S., Tyler I.M., 2014. A Magnetotelluric Traverse Across the Eastern Part of the Capricorn Orogen : Geological Survey of Western Australia, Report 135, 49 pp. Egbert G.D., Kelbert A., 2012. Computational recipes for electromagnetic inverse problems. Geophys. J. Int. , 189( 1), 251– 267. https://doi.org/10.1111/j.1365-246X.2011.05347.x Google Scholar CrossRef Search ADS   Groom R.W., Bailey R.C., 1989. Decomposition of magnetotelluric impedance tensors in the presence of local three-dimensional galvanic distortion, J. geophys. Res. , 94( B2), 1913– 1925. https://doi.org/10.1029/JB094iB02p01913 Google Scholar CrossRef Search ADS   Groom R.W., Bailey R.C., 1991. Analytic investigations of the effects of near-surface three-dimensional galvanic scatterers on MT tensor decompositions, Geophysics , 56( 4), 496– 518. https://doi.org/10.1190/1.1443066 Google Scholar CrossRef Search ADS   Heise W., Pous J., 2003. Anomalous phases exceeding 90° in magnetotellurics: anisotropic model studies and a field example, Geophys. J. Int. , 155( 1), 308– 318. https://doi.org/10.1046/j.1365-246X.2003.02050.x Google Scholar CrossRef Search ADS   Ichihara H., Mogi T., 2009. A realistic 3-D resistivity model explaining anomalous large magnetotelluric phases: the L-shaped conductor model, Geophys. J. Int. , 179( 1), 14– 17. https://doi.org/10.1111/j.1365-246X.2009.04310.x Google Scholar CrossRef Search ADS   Ichihara H., Mogi T., Yamaya Y., 2013. Three-dimensional resistivity modelling of a seismogenic area in an oblique subduction zone in the western Kurile arc: constraints from anomalous magnetotelluric phases, Tectonophysics , 603, 114– 122. https://doi.org/10.1016/j.tecto.2013.05.020 Google Scholar CrossRef Search ADS   Jones A.G., 1999. Imaging the continental upper mantle using electromagnetic methods, Lithos , 48( 1-4), 57– 80. https://doi.org/10.1016/S0024-4937(99)00022-5 Google Scholar CrossRef Search ADS   Jones A.G., Groom R.W., Kurtz R.D., 1993. Decomposition and modelling of the BC87 dataset, J. Geomagn. Geoelec. , 45( 9), 1127– 1150. https://doi.org/10.5636/jgg.45.1127 Google Scholar CrossRef Search ADS   Kelbert A., Meqbel N., Egbert G.D., Tandon K., 2014. ModEM: a modular system for inversion of electromagnetic geophysical data, Comput. Geosci. , 66, 40– 53. https://doi.org/10.1016/j.cageo.2014.01.010 Google Scholar CrossRef Search ADS   Krieger L., Peacock J.R., 2014. MTpy: a Python toolbox for magnetotellurics, Comput. Geosci. , 72, 167– 175. https://doi.org/10.1016/j.cageo.2014.07.013 Google Scholar CrossRef Search ADS   Ledo J., Queralt P., Martí A., Jones A.G., 2002. Two-dimensional interpretation of three-dimensional magnetotelluric data: an example of limitations and resolution, Geophys. J. Int. , 150( 1), 127– 139. https://doi.org/10.1046/j.1365-246X.2002.01705.x Google Scholar CrossRef Search ADS   Lezaeta P., Haak V., 2003. Beyond magnetotelluric decomposition: induction, current channeling, and magnetotelluric phases over 90°, J. geophys. Res. , 108( B6), 2305, doi:10.1029/2001JB000990. https://doi.org/10.1029/2001JB000990 Google Scholar CrossRef Search ADS   Lilley F.E.M., Weaver J.T., 2010. Phases greater than 90° in MT data: analysis using dimensionality tools, J. appl. geophys. , 70( 1), 9– 16. https://doi.org/10.1016/j.jappgeo.2009.08.007 Google Scholar CrossRef Search ADS   Martí A., Queralt P., Ledo J., 2009. WALDIM: A code for the dimensionality analysis of magnetotelluric data using the rotational invariants of the magnetotelluric tensor, Comput. Geosci. , 35( 12), 2295– 2303. https://doi.org/10.1016/j.cageo.2009.03.004 Google Scholar CrossRef Search ADS   Martí A., 2014. The role of electrical anisotropy in magnetotelluric responses: from modelling and dimensionality analysis to inversion and interpretation, Surv. Geophys. , 35( 1), 179– 218. https://doi.org/10.1007/s10712-013-9233-3 Google Scholar CrossRef Search ADS   Mccuaig C., Hronsky, J., 2014. The mineral system concept: the key to exploration targeting, pp. 153– 175, BT–Building Exploration Capability for, Society of Economic Geologists, Inc. Pek J., 2009. Effects of electrical anisotropy upon magnetotelluric data: modelling and experiments, in Modern Methods of Electromagnetic Data Measurement, Processing and Interpretation , pp. 110– 135, ed. Spichak V.V., Librokom Publ (in Russian). Pous J., Heise W., Schnegg P., Muñoz G., Martí J., Soriano C., 2002. Magnetotelluric study of the Las Cañadas caldera (Tenerife, Canary Islands): structural and hydrogeological implications, Earth planet. Sci. Lett. , 204( 1–2), 249– 263. https://doi.org/10.1016/S0012-821X(02)00956-1 Google Scholar CrossRef Search ADS   Selway K., Sheppard S., Thorne A., Johnson S., Groenewald P., 2009. Identifying the lithospheric structure of a Precambrian Orogen using magnetotellurics: the Capricorn Orogen, Western Australia, Precambrian Res. , 168( 3–4), 185– 196. https://doi.org/10.1016/j.precamres.2008.09.010 Google Scholar CrossRef Search ADS   Selway K., Thiel S., Key K., 2012. A simple 2-D explanation for negative phases in TE magnetotelluric data, Geophys. J. Int. , 188( 3), 945– 958. https://doi.org/10.1111/j.1365-246X.2011.05312.x Google Scholar CrossRef Search ADS   Sheppard S., Fletcher I.R., Rasmussen B., Zi J., Muhling J.R., Occhipinti S.A., Wingate M.T.D., Johnson S.P., 2016. A new Paleoproterozoic tectonic history of the eastern Capricorn Orogen, Western Australia, revealed by U–Pb zircon dating of micro-tuffs, Precambrian Res. , 286, 1– 19. https://doi.org/10.1016/j.precamres.2016.09.026 Google Scholar CrossRef Search ADS   Simpson F., Bahr K., 2005. Practical Magnetotellurics , Cambridge Univ. Press. Google Scholar CrossRef Search ADS   Thiel S., Heinson G., Gray D.R., Gregory R.T., 2009. Ophiolite emplacement in NE Oman: constraints from magnetotelluric sounding, Geophys. J. Int. , 176( 3), 753– 766. https://doi.org/10.1111/j.1365-246X.2008.04053.x Google Scholar CrossRef Search ADS   Weaver J.T., Agarwal A.K., Lilley F.E.M., 2000. Characterization of the magnetotelluric tensor in terms of its invariants, Geophys. J. Int. , 141( 2), 321– 336. https://doi.org/10.1046/j.1365-246x.2000.00089.x Google Scholar CrossRef Search ADS   Weckmann U., Ritter O., Haak V., 2003. A magnetotelluric study of the Damara Belt in Namibia, Phys. Earth planet. Inter. , 138( 2), 91– 112. https://doi.org/10.1016/S0031-9201(03)00079-7 Google Scholar CrossRef Search ADS   SUPPORTING INFORMATION Supplementary data are available at GJI online. Figure S1. Example of sites with anomalous phases recorded in the study area (dots and triangles) and its comparison with the ongoing 3-D inversion models (lines) for both, off-diagonal (left) and diagonal (right) components. Figure S2. Phase tensor ellipses (Phi.min) and induction arrows (black: real; red: imaginary) for each BBMT site at six different periods. Figure S2. 3-D synthetic models used to try and explain the observed POQ and calculated phases curves for some sites with anomalous YX phases. Dots and triangles: observed phases; Lines: Synthetic phases. (a) General background model (Model S) for 0–8.4 km depth; at greater depths the model is a homogeneous half-space of 10000 Ωm. (b) Set of models with a regional E–W trending conductive body at depth (8.4–9.8 km; resistivity: 2 Ωm). (c) Model S-21, same structure than Model S-7 with a local conductive body of 10 Ωm in the centre of the Marymia Inlier from 0.5 to 8.4 km depth. (d) Set of models performed modifying the thickness of the basins to the east. The southernmost basin is closed at depths greater than 1.8 km. Model S-3 is equivalent to Model S, without regional E–W conductive body; Model S3.7 is equivalent to Model S-7, with the same regional E–W conductive body. (e) Tests performed to study the effect of the resistivity distribution in the Palaeoproterozoic basin in Model S-7, since resistivities larger than 50 Ωm are expected below 2 km depth. Layered basins with resistivity values increasing with depth provide similar results in terms of anomalous phases at long periods. Please note: Oxford University Press is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the paper. © The Author(s) 2017. Published by Oxford University Press on behalf of The Royal Astronomical Society.

Journal

Geophysical Journal InternationalOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off