Influence of Tectonic and Geological Structure on GIC in Southern South Island, New Zealand

As part of a 5‐year project to assess the risk posed by geomagnetically induced currents (GIC) to the New Zealand electrical transmission network, long‐period magnetotelluric (MT) measurements have been made at 62 sites in southern South Island of New Zealand, a region where there was an absence of previous MT data. The data are largely 3‐dimensional in character, but show distinct features that can be related to the known tectonic and geological structure. In this work we focus on how the measured MT impedance tensors, and a simple interpretation of conductivity structure, can be used to assess the influence of tectonic and geological structure on GIC. We use the impedance tensors to calculate the magnitudes and orientations of induced electric fields in response to various orientations of inducing magnetic field. The electric fields so calculated are then used in a simplified model of the transmission network to calculate GIC at grounded substations. Our results confirm that tectonic/geological structure in the lower South Island and the resulting electrical conductivity variations have important impacts on the GIC magnitude. In the south‐west, smaller induced electric fields, associated with the higher conductivity in that region, lead to much reduced GIC at a substation in that area. In contrast, higher electric fields occurring in a NW‐SE band across the center of the region, contribute to much larger GIC in Dunedin city. Our results thus help explain the observed GIC reported at transformers in the region.

Plain Language Summary Variations in the Earth's magnetic field during magnetic storms produce induced currents in the ground which in certain circumstances may present a risk to an electricity transmission network.Understanding the risk in any given region requires knowledge of the local ground electrical conductivity structure.To help map the structure across southern South Island, New Zealand, we have made long period magnetotelluric measurements at 62 sites.We use these measurements to calculate the electric fields which would be induced in the ground due to magnetic field variations.Using a simplified representation of the electrical transmission network, we use these calculated electric fields to assess how particular features of the geological structure influence the currents (geomagnetically induced currents-GIC) that can be produced in the power network.

INGHAM ET AL.
INGHAM ET AL.

10.1029/2023SW003550
2 of 17 orientation of induced electric fields across a region is to calculate them directly from the magnetic field spectrum for a specified geomagnetic storm using magnetotelluric (MT) impedance tensors from a two-dimensional distribution of sites.Prior to 2020, MT sites in the South Island have largely been limited to transects of sites across the northern and central parts of the Island (Ingham, 1996;Wannamaker et al., 2002Wannamaker et al., , 2009) ) making the use of MT data in assessing GIC risk impractical.As a step to improving coverage of MT data, an extensive long period MT survey across Otago and Southland in the southern part of the South Island of New Zealand has recently been completed.Full details of analysis and inversion modeling of this data to reveal details of the lithospheric conductivity structure will be described in a future publication.In this paper we focus on how the measured MT impedance tensors, and a simple interpretation of conductivity structure, can be used to assess the influence of tectonic and geological structure on GIC in the southern part of the South Island of New Zealand.
We start by discussing the tectonic and geological framework of the study area and the extent of the MT data survey.We then use apparent resistivity and phases calculated from the determinant impedance to give a broad picture of conductivity variations across the area.We use the impedance tensors to calculate the magnitudes and orientations of induced electric fields in response to various orientations of a uniform inducing magnetic field.The calculated electric fields are then used in a simplified electrical model of the transmission network in the region to calculate GIC at grounded substations.By replacing anomalously large or small geoelectric fields associated with prominent tectonic/geological features with more uniform values, we assess how the tectonic/ geological structure influences calculated GIC.Carried out in the frequency domain, our analysis differs from that employed by Bedrosian and Love (2015) who used impedance tensors measured as part of the Earthscope program to map the time variation of induced electric fields across the mid-western United States, and related the magnitude of the induced fields to the underlying conductivity structure.Similar investigations were also performed for the north of England and southern Scotland by MacKay and Whaler (2006), who also looked at the influence of galvanic distortion on the orientation and magnitude of electric fields, by Nakamura et al. (2018) in a study of GIC in the Japanese 500-kV power grid, and by Marshalko et al. (2020) for the eastern United States.

Geological/Tectonic Setting and Previous Geophysical Measurements
New Zealand lies on the boundary between the Australian and Pacific tectonic plates.Throughout the South Island this boundary is marked by the Alpine Fault, a transform fault which, over the last 25 Myr, has seen about 480 km of lateral movement.The convergence of the Pacific and Australian plates along the fault has led to the uplift of the Southern Alps along the central part of the South Island.The geological structure of the study area (shown in Figure 1) reflects both this movement and the complex history of accretion of terranes and periodic rifting events (King, 2000) at the margin of Gondwana from around 300 Myr (Sutherland et al., 2000).As a result, as described by Mortimer et al. (2002), the basement across the study area is made up of a "collage" of volcano-sedimentary terranes of which some have been affected by metamorphism and/or intrusion.
From the west coast of Fiordland to Invercargill and the Hollyford Fault the basement consists of the 250-100 Myr Median Batholith and the adjacent Brook Street Terrane.Both represent volcanic arcs generated behind subduction zones.East of these features, the Murihiku Terrane (MuT) is composed primarily of sandstones and siltstones, and is bounded on its north-east by the Foothills Fault which delineates the south-western boundary of the Dun Mountain-Matai Terrane.The Dun Mountain-Matai Terrane is covered by Cenozoic and Cretaceous sediments along much of its length and consists of obducted oceanic crust (the Dun Mountain Ophiolite Belt (DMOB)) overlain by continental margin sandstones and limestones.The DMOB can be traced for about 1,000 km into the North Island through the Stokes Magnetic Anomaly (Hunt, 1978).To the north are Permian and Triassic greywackes which over a large part of the area have been metamorphically altered to schist (the Haast schist) which is comprised of fault controlled north-east trending parallel ridges and basins (Ballance, 2009).
The majority of the study area is overlain by Cenozoic and Cretaceous sedimentary deposits, river gravels and alluvium.The sediments are thickest across the eastern portion of the Median Batholith between Fiordland and the western edge of Southland, in a region known as the Moonlight Tectonic Zone-MTZ- (Ballance, 2009).This marks the location of a spreading center between the Pacific and Australian tectonic plates, active between 45 and 25 Myr (King, 2000).The only recent volcanism in the study area is related to the 13-10 Myr Dunedin volcano, the remnants of which now make up the Dunedin peninsula and are unrelated to the accretionary and rifting events.

Magnetotelluric Measurements
Long period magnetotellurics (MT) is a passive geophysical method that incorporates a cartesian layout of two electric dipoles and a fluxgate magnetometer to record Earth's naturally occurring electric and magnetic fields.These measurements are used to calculate an impedance tensor relating variations in the induced electric field to those in the magnetic field.The impedance tensor has components Z xx , Z xy , Z yx , and Z yy and relates the induced horizontal components of the electric field to the variations in the horizontal components of the magnetic field through  = ( + )∕0 (1) is a function of frequency and depends upon the electrical conductivity structure of the earth.Long period MT refers to data acquisition within the frequency range of 0.0001-1 Hz, for which the source signal originates from the interaction of the solar wind with the geomagnetic field.
The locations of 62 long period magnetotelluric sites across Otago and Southland are shown in Figure 1.The MT measurements were made using LEMI-417M instrumentation produced by the Lviv Centre of the Institute for Space Research, Ukraine, utilizing fluxgate magnetometers with a frequency band from 0.3 to 10 −5 Hz, and Pb/ PbCl 2 electrodes constructed in-house by GNS Science.Measurements were made over a period of 15 months from February 2021 to May 2022, with eight sites at a time operating simultaneously with a remote reference site situated in the center of the North Island.At each site data were recorded for approximately 1 month with a sampling interval of 1 s.Data quality was generally good with the exception of three sites where electric or magnetic field cables were disturbed by wildlife.Impedance tensor estimates were calculated, in geographic coordinates (x-axis oriented to north, y to east) in the period range from 5 to 100,000 s.Examples of the calculated Analysis both of phase tensors (Bibby et al., 2005;Caldwell et al., 2004) and of dimensionality indices (Marti et al., 2009;Weaver et al., 2000) suggest that the overall electrical conductivity structure of this part of the South Island is, as might be expected from the tectonics, highly complex, and will require 3-dimensional numerical modeling/inversion.As our interest in the present paper is to analyze the effects of tectonic/geological structure on induced electric fields, and hence GIC, we use pseudosections of the so-called determinant apparent resistivity and phase measurements to gain an initial understanding of the principal features of the conductivity structure.

Determinant Apparent Resistivity and Phase
Although, as indicated above, complete understanding of the conductivity structure associated with the study area requires full 3-dimensional inversion modeling, some simpler methods of gaining an insight into the principal structural features are available.In particular the use of parameters which are derived from the magnetotelluric impedance tensor and are invariant upon rotation suggest the possibility of assessing the variation in conductivity structure with depth beneath a site.Ingham (1988) investigated the use of the apparent resistivity and phase calculated from the determinant impedance defined by which has the advantage of being a function of all four elements of the impedance tensor.If the true impedance tensor is affected by galvanic distortion through a real distortion tensor then the measured determinant is related to it only through a multiplicative constant  √  −  .Hence the determinant phase is unaffected and the apparent resistivity only affected through a possible static-shift.
To identify changes in conductivity structure across the study area pseudosections of the apparent resistivity and phase calculated from Z det have been plotted for four different transects as marked in Figure 1.Transect AA′ runs approximately NE to SW and includes four sites on the schist terrain, one just to the south-west of the DMOB, and 4 sites lying in the Moonlight Tectonic Zone.Transect BB′ is to the south-east of this and includes sites which lie in/on all of the major geological regions-the schist, the DMOB, the MuT and the Median Batholith/Brook Street Terrane.The two transects CC′ and DD′, further to the east both have five sites on the schist, one on the DMOB, and two to the south on the MuT.
Figure 3 shows the determinant apparent resistivity and phase pseudosections for all four transects.For transect AA′ the phase pseudosection shows low phases (blue to purple) persisting to longer period (lower frequency) beneath the MuT and the MTZ compared to sites on the schist.The main feature in the determinant apparent resistivity pseudosection is the large contrast between sites in the MTZ and those to the north.Although the smooth variation in phase with location suggests that the higher ρ a values at some sites, especially 134 and 161, may be due to static-shift, it is clear that the MTZ is marked by much lower apparent resistivities.Commensurate with this, due to the much smaller skin-depth in a higher conductivity, it is likely that structure seen beneath this zone is at a much shallower depth than at sites on adjacent more resistive terrain.There is also a rise in phase values at the shortest periods beneath site 153.
The pseudosections for transect BB′ show similar features across the schist and the DMOB, with the same extension of low phase to longer periods at the more south-western sites.Most noticeable is the very low ρ a values at site 159.This site lies on the Cenozoic sediments which cover the MuT and the apparent resistivity shows a very large contrast with values at the adjacent site 167 which lies on the Brook Street Terrane immediately to the east of the inferred location of the MTZ.It is possible that this reflects a connection between the MTZ and this portion of the Murihiku.
Pseudosections for transects CC′ and DD′ are also shown in Figure 3 and are remarkably similar.At the shortest periods there is a similar decrease in phase from north-east to south-west as seen on transects AA′ and BB′.However, there is also a very noticeable change in phase at longer periods across the Dun Mountain Ophiolite Belt with markedly lower values in the south-west compared to the north-east.On both transects the single site within the DMOB also shows slightly higher ρ a values which, given the observed change in phase, are likely not an effect of static-shift.Although there are relatively large gaps between these sites (140 and 138) and adjacent sites to the north (131 and 130), the determinant apparent resistivity and phase suggest that there is actually a band of high resistivity to the north of the DMOB.
INGHAM ET AL.

10.1029/2023SW003550
6 of 17 These features can also be clearly seen in Figure 4 which shows contour plots of log10 apparent resistivity and phase across the study area at three different periods of variation.In particular the lower apparent resistivity associated with the MTZ, extending east into MuT, is very clearly evident at all three periods.Also evident at periods of 31.6 and 316 s is the lower phase beneath the Murihiku compared to the schist which is seen on all 4 transects in Figure 3.At the two longer periods phases higher than 45° are also seen right across the schist terrane.Higher resistivity immediately to the north of the Dun Mountain Ophiolite Belt is less obvious but can also be seen.
In summary, analysis of the determinant pseudosections, supported by individual apparent resistivity curves calculated from the impedance tensors, suggests the following: 1.The conductivity structure in the western part of the study area where the DMOB trends S-N and is adjacent to the MTZ, and to the south of this where the Murihiku abuts the Median Batholith, is complex and clearly 3-dimensional.2. In this region, the MTZ marks an area of significantly lower resistivity, certainly in the near-surface but possibly to considerable depth.This is in contrast to the adjacent higher resistivity Median Batholith to the west.3.There are indications that parts of the MuT to the east of the MTZ may also have lower resistivity than areas to the north and south.4. Across the schist of Central Otago the structure appears relatively uniform.Phase values greater than 45° at periods longer than a few hundred seconds suggesting an increase in conductivity at depth. 5.In the south-eastern part of the study area there is evidence for a change in conductivity structure across the DMOB.The DMOB itself, and possibly the schist immediately to the north may exhibit higher resistivity than occurs either further to the north or on the MuT to the south.

Induced Electric Fields
To investigate any possible influence of the geological and tectonic structure on the production of GIC in this part of the transmission network, it is assumed that magnetic field variations are the same across the entire study area.The validity of using a spatially uniform magnetic field variation was tested by Divett et al. (2020), who found that it gave only minor differences in calculated geoelectric fields when compared to a spatially varying field.Any magnitude and orientation of magnetic field variation at some period can be applied to Equations 1 and 2 and used to calculate the magnitude and orientation of the resulting induced electric field at an MT site.These induced electric fields, when integrated over the topology of the transmission network, result in the production of GIC.
As a starting point to understand the level of influence of geological structure on the production of GIC in this region, in Figure 5 we show the amplitude and orientations of electric fields produced in Southland/Otago by 100 nT magnetic field variations in north, east, north-west and north-east orientations for a period of 30 srepresentative of the shortest period of variation for which data are available.The upper parts of Figure 5 show the magnitudes of the induced electric fields and there are two main features which are evident, and which reflect the contour plots of apparent resistivity shown in Figure 4.For all orientations of inducing field the low resis tivity found in the Moonlight Tectonic Zone and across the western part of the MuT (Figure 4) results in much smaller magnitude electric fields than elsewhere in the study area.Especially for eastward and north-westward orientations of the inducing field, but also apparent for a northward inducing field, there is a relatively broad zone immediately to the north of the DMOB where the induced electric fields are significantly higher.This coincides with sites (e.g., 134, 135, 138 and 140) where the determinant apparent resistivity (Figure 3) is higher than either to the north or the south.Although, for north, north-east and north-west inducing fields, significantly higher electric fields are found to be induced at a single site close to the north-west end of the DMOB, these are likely to be the result of static-shift of the apparent resistivity as suggested by the much higher apparent resistivity at this site which is seen in Figure 4 but is not associated with any anomalous phase.Apart from these two broad features, 9 of 17 higher induced electric fields also occur both immediately inland from Dunedin, and to the north and north-west of Invercargill, the latter corresponding to the exposed portion of the Brook Street Terrane.Significantly larger induced fields also occur on, at least, the eastern edge of the Median Batholith where it borders the Moonlight Tectonic Zone.For each orientation of inducing magnetic field, the lower parts of Figure 5 show that, although there is some degree of scatter and a few anomalous sites, the orientations of the principal axes of the electric field ellipses is, in each case, roughly perpendicular to the orientation of the inducing field.Thus, a northward inducing field produces electric field axes which are broadly aligned east-west.It can also be noted that for north and north-east inducing fields there is a degree of agreement in the orientation of electric field axes with the orientation of the transmission lines connecting the Dunedin substations (HWB and SDN, collectively referred to as DUN) with both ROX and INV.To a large degree these transmission lines also cross the regions where the induced electric fields are larger, particularly the band of higher electric field magnitudes to the north of the DMOB.For a north-west inducing field the electric field principal axes are also broadly parallel to the transmission line from HWB/SDN to INV and are also approximately parallel to that connecting INV to ROX, although closer to INV the latter crosses the region of reduced electric field magnitude associated with the MTZ and the MuT.It is only for a north-east inducing field that electric field axes align with the transmission line from INV to MAN, and for the majority of its length this passes through the region of significantly smaller induced electric fields.
Although the smaller rate of change of magnetic field associated with longer period variations results in smaller magnitude induced electric fields, the patterns of induced fields shown in Figure 5 for a variation of period 30 s, are similar to those found for other periods of variation.This is consistent with what is seen in Figure 4 which shows the similarity between determinant apparent resistivity and phase plots across two decades of period.

A Simplified Electrical Model
To better assess the impact that the tectonic and geological structure, especially that associated with the MTZ/ MuT and the DMOB, have on GIC we use a simplified calculation of currents in the network connecting the MAN, INV, ROX and Dunedin substations.Although there are multiple lines connecting some of these substations, for simplicity each connection is represented as a single line, with lengths and orientations which are approximated as listed in Table 1.With this simplification the network can broadly be represented by the circuit shown in Figure 6.In this circuit each substation is shown as earthed with a resistance R s which represents the sum of the transformer resistance and the actual grounding resistance.Each line between substations is shown as having a resistance and a voltage source.Thus, R I-M is the line resistance between INV and MAN, and V I-M is a voltage source representing the potential difference produced between the two substations-that is, the line integral of the electric field along the transmission line.Following Divett et al. (2017) we take R s as 1.1 Ω and the line resistances as 0.05 Ω/km.If the currents (GIC) at each substation and in each line are represented as shown, as detailed in Appendix A, Kirchhoff's Laws can be applied to calculate the resulting currents from the matrix equation Table 1 Lengths and Orientations of Transmission Lines as Represented in Figure 6 INGHAM ET AL. 10.1029/2023SW003550 12 of 17 where represents the length of transmission line in a particular grid cell, and the vector electric field in that cell.The scalar product resolves the field in the cell onto the orientation of the transmission line.Summation over all the grid cells that the transmission line passes through then gives the potential difference.
The influence of specific tectonic/geological features on GIC can be assessed by omitting particular structural features as shown on the maps in Figure 4. Specifically, the effect of the MTZ and Murihiku on the production of GIC can be tested by replacing the low electric field regions in cells B, G, and H associated with the MTZ/MuT with fields of 1,000 mV/km, a value which represents the approximate average of the electric field magnitude at all the MT sites, and is found to be independent of the orientation of the inducing field.Given the relatively large size of the grid cells the manner in which to test the impact of the higher electric fields to the north of the DMOB is less clear.Given this limitation this has been attempted by replacing the broad band of higher (≥1,000 mV/km) electric fields north of the DMOB (cells C, D, E, F, J and K) by fields of 800 mV/km.This, by necessity, includes a significant portion of the south-western part of the schist terrane, and, indeed, means that for a north-east inducing field the assumed electric field in cell D is actually increased (see Appendix A).As a result of these limitations any clear effect of removing the DMOB is probably minimized.In any event, although calculations using such a simplified model are necessarily an approximation, the results of such an analysis do prove useful in investigating the manner in which these structural features effect the magnitudes of GIC at the four substations.

Results
The results of calculations for north, north-east and north-west oriented 100 nT inducing magnetic fields, based on the magnitudes of induced electric fields calculated for a period of 30 s, are shown graphically in Figure 8.The calculation of currents for the full structure (green bars in Figure 8), show that GIC at DUN are opposite in sign to GIC at MAN for all orientations of the inducing magnetic field (Figures 8a, 8c, and 8e).This is unsurprising given that DUN and MAN can be regarded as being at the eastern and western ends of the transmission line network respectively.It can also be seen that the sense of GIC at both INV and ROX depends on the orientation of the inducing field, reflecting the fact that these two substations can be regarded as intermediate between DUN and MAN, with the sense of GIC heavily dependent on the distribution of currents in the transmission lines.
As noted above, large GIC are commonly seen at the DUN substations.The overall pattern of calculated GIC seen in Figure 8 suggests that the origin of these large GIC differs depending on the orientation of the inducing magnetic field.For a northward inducing field the large GIC at DUN are associated with oppositely oriented GIC at both INV and MAN, with only very small GIC at ROX.However, for a north-east oriented inducing field the large negative GIC at DUN are related to significant positive GIC at ROX and MAN, with only very small GIC at INV.With a north-west inducing field orientation GIC at ROX are of the same sense as at DUN and are balanced by opposite GIC at both INV and MAN.
Removal of the low electric fields seen in the MTZ and MuT (red bars in Figure 8) leads to a significant increase in current in the transmission line between INV and MAN for both north and north-east oriented inducing fields (Figures 8b and 8d).In both cases this leads to an increase in GIC at MAN of near 100% (Figures 8a and 8c).
For a northward inducing field removal of the MTZ leads to a decrease in GIC at INV of ∼5 A (Figure 8a).As, for this inducing field, GIC at INV have the same sense as GIC at MAN this implies that the increase in GIC at MAN, and the increased INV-MAN current, come not solely from the reduced GIC at INV but are also due to the increase in currents in both the DUN-INV and ROX-INV transmission lines (Figure 8b).In contrast, for a north-eastward inducing field GIC at INV have the opposite sign to those at MAN.Thus, the very large increase in current in the INV-MAN line (Figure 8d) can be seen to be primarily a result of current flowing into the line at INV and out at MAN.For a northward inducing field GIC at ROX are very small and appear independent of the MTZ structure.While GIC at ROX are larger for a north-east inducing field (Figure 8c) they are similarly essentially independent of the presence of absence of the low electric fields in the MTZ/Murihiku.For a north-west inducing field (Figures 8e and 8f) the fact that the INV-MAN transmission line is near parallel to this orientation means that, although there is some redistribution of both GIC and currents in the transmission lines, including a reversal of GIC at ROX, the overall effect of removing the MTZ is very small.
Although it appears that the presence of the region of low resistivity and small induced electric fields associated with the MTZ are responsible for the relatively small GIC seen at MAN, the impact of the DMOB is much less clear.As indicated this probably partly reflects the difficulty in representing the region of enhanced electric fields on the chosen 0.5° scale, but, nonetheless, some inferences can be drawn from Figure 8.In particular it can be seen (Figures 8a and 8e) that, for inducing fields oriented north and north-west, removing the high electric field values associated with cells C, D, E, F, J and K leads to a reduction in the size of GIC at both DUN and INV.This is entirely consistent with an associated reduction in current in the DUN-INV transmission line seen in Figures 8b  and 8f.At DUN the drop in GIC is between 4 and 6 A, for both orientations of inducing field about 20% of the value calculated with the full structure, whilst at INV the reductions in GIC are closer to 25%.At ROX GIC are negative (i.e., into the substation, Figure 6) for both these orientations of inducing field.There is very little change in currents for a north-east oriented inducing field.Again, this probably reflects the relationship between the orientation of the inducing field and the transmission lines.The north-east inducing field orientation is very close to that of the ROX-INV transmission line and leads to only a very small potential difference between ROX and INV, while there is a much reduced potential difference between DUN and INV compared to the other two orientations of inducing field.

Conclusion
In summary, although the limitations of such a simplified calculation must be emphasized, the results suggest that the tectonic/geological structure in southern South Island, and the way it has resulted in electrical conductivity variations, does have a impact on GIC.Most obvious and significant is that smaller induced electric fields, associated with the higher conductivity of the Moonlight Tectonic Zone and the MuT, clearly have the effect of reducing the GIC observed at MAN.Without this zone of significantly reduced electric fields any inducing field with a predominantly north to north-east orientation (which encompasses local magnetic north) would result in much larger GIC at MAN.The resultant increased current between INV and MAN would also have an impact on other substations.
For north and north-west oriented inducing magnetic field, the higher induced electric fields associated with DMOB, and to its north, appears to be a contributor to the large GIC which are observed at the Dunedin substations.Although the calculated effect of removing the DMOB as described is gives a reduction of ∼5 A at DUN, the very broad manner in which the DMOB has been represented in this calculation likely means that its true contribution to the large GIC at DUN is greater.Adjustments to the transmission line model (Figure 6) such as including bends in the lines are also likely to have only a secondary effect on the calculated currents.This is consistent with Divett et al. (2020) who found that assumption of linearity of transmission lines made little difference in the actual calculation of GIC.
Both the magnitude and sense of GIC at both INV and ROX are highly dependent on the orientation of the inducing field.At INV the orientation of induced electric field relative to the transmission lines from the other three substations is important in determining the significance of both the MTZ and the DMOB.At ROX neither has much impact on GIC.
GIC produced in a power network result from the integral of the electric field, induced by magnetic activity, along the length of the powerlines.The magnitude of induced electric fields is dependent not only on the rate of change of the magnetic field, but also on the electrical conductivity of the ground.The orientation of the induced fields is similarly dependent not only on the orientation of the inducing magnetic field, but on the spatial distribution of conductivity.Thus, in any region that is susceptible to the generation of GIC in transmission lines, a knowledge of the spatial distribution of magnitudes and orientations of the induced electric field is important for understanding GIC.This work thus emphasizes the importance of improved ground conductivity measurements in GIC-focused Space Weather studies.It also helps explain the very high GIC regularly reported in Dunedin, and why the transformers in that city has been identified as particularly high risk during an extreme space weather event (Mac Manus, Rodger, Dalzell, et al., 2022).
INGHAM ET AL.
which, with assumed values for R s , the resistance per km of the transmission lines, and estimated values for the voltages, may be solved for the eight currents.

A2. Magnitudes and Orientations of Induced Electric Fields
The tables below show the average values of the magnitude and orientation of the induced electric field in each grid cell (as shown in Figure 7) for north, north-east and north-west orientations of the inducing magnetic field for a period of 30 s (Tables A1, A2, and A3).
passes off-land in the north-west of the study area and immediately to the south-west of the lower South Island the Australian Plate subducts steeply under the Pacific Plate at the Puysegur Trench.

Figure 1 .
Figure 1.Generalized geology and tectonic structure of the study area and location of MT sites.Dotted lines mark the pseudosection transects shown in Figure 3.

Figure 3 .
Figure 3. Pseudosections of log10 apparent resistivity (Rhoa) and phase calculated from the determinant impedance for transects AA' to DD' as marked in Figure 1.

Figure 4 .
Figure 4. Contour plots of log10 apparent resistivity (Rhoa) and phase calculated from the determinant impedance for three periods of variation.

Figure 5 .
Figure 5. Magnitudes and principal axes of electric fields induced by a 100 nT variation in the magnetic field in north, east, north-west and north-east orientations.The locations of substations mentioned in the text are also shown, and the locations of high-voltage power transmission lines are shown in blue.

Figure 8 .
Figure 8. Calculated currents (geomagnetically induced currents) at substations and in transmission lines for three orientations of the inducing magnetic field as indicated.Green bars-currents calculated when including the effect of all geological structures; red bars-currents calculated with the effect of the Moonlight tectonic Zone removed; blue bars-currents calculated with the effect of the Dun Mountain Ophiolite Belt removed.

Table 2
Average Values of the Magnitude and Orientation of Induced Electric Fields in Each Grid Cell for a Northward Inducing Field of Magnitude 100 nT at Period 30 s

Table A1 Average
Values of the Magnitude and Orientation of Induced Electric Fields in Each Grid Cell for a Northward Inducing Field of Magnitude 100 nT at Period 30 s

Table A2 Average
Values of the Magnitude and Orientation of Induced Electric Fields in Each Grid Cell for a North-EastwardInducing Field of Magnitude 100 nT at Period 30 s

Table A3 Average
Values of the Magnitude and Orientation of Induced Electric Fields in Each Grid Cell for a North-WestwardInducing Field of Magnitude 100 nT at Period 30 s