Mark A. Clilverd

Geomagnetically Induced Currents (GICs) are a key space weather hazard to ground‐based infrastructure, and can cause mis‐operation or even equipment failure. The solar wind interacts with the Earth's geomagnetic field, causing the magnetic field at ground level to vary with time. This variability causes GICs in grounded, conducting networks via coupling with the subsurface geology. Most studies of the hazard focus on short, high amplitude GICs, but long intervals of lower amplitude GIC can also be problematic, representing a rare but known pathway to equipment damage and power failure. We evaluate an interval in June 2015 where the power network in New Zealand experienced long‐duration, steadily increasing GICs across the South Island. These GICs were up to 20A and lasted approximately 90 min during an otherwise moderate storm interval, while New Zealand was in the dawn sector. We investigate the magnetospheric cause, attributing them to a possible diversion of the ring current into the ionosphere. We test the ability of modeling and proxies to capture this hazard, finding that it can be recreated, but is dependent upon the regional geology, and requires detailed knowledge of the network configuration. We model two hypothetical examples, (a) where the magnetic perturbation is larger and (b) where a different mid‐latitude location is exposed, finding that larger currents are possible and that the vulnerability is not unique to New Zealand. Our case study highlights the limitations of geomagnetic indices, which we show to be vulnerable to influence from such rare mid‐latitude phenomena.

Geomagnetic storms represent a space weather hazard to power transmission networks due to the effects of induced geo‐electric fields within the conducting surface of the Earth. These drive electric currents in power transmission lines which can flow to ground through the neutral‐ground connections of transformers. Geomagnetically induced currents (GIC) can negatively impact the operation of high voltage transformers through asymmetric half‐cycle transformer core saturation. In this study GIC measurements, derived effective current, and reactive power responses ( Q ) of the single‐phase bank autotransformer, T4, operating at 220 kV in the Halfway Bush substation, Dunedin, New Zealand have been analyzed over the period 2013–2017. During 8 elevated GIC events linear enhancements of reactive power consumption ( Q _con , MVAr) occurred, even with comparatively low levels of GIC (i.e., <7 A). This is consistent with transformer core saturation where there is little tolerance or “headroom” to GIC in the design for single‐phase bank transformers. Reactive power measurements show high variability. The removal of the non‐GIC variability is difficult and introduces uncertainty into the identification of GIC‐driven responses. In this analysis we consider two techniques to isolate GIC‐induced reactive power responses. We find that transformer T4 exhibited a reactive power response of ∼0.115 MVAr/A. The results are a factor of 2–3 larger than MVAr changes in a nearby three‐phase, three‐limb autotransformer, aligning with previous research. The results are consistent with, but less extreme than modeling studies found for transformer units operating at significantly higher voltages.

The geomagnetic storm of 10–11 May 2024 (G5‐level) caused major ionospheric disturbances, providing an opportunity to examine its wide‐ranging effects. This study analyzes ionospheric perturbations over mid‐ and high‐latitude regions using Very Low Frequency (VLF) radio signal observations and TIMED/SABER satellite data in both hemispheres. Significant VLF amplitude anomalies were recorded along five paths (NAA‐BRN, NAA‐ASB, NAA‐KIL, NAA‐St. Johns, and NAA‐Rothera), indicating strong lower ionospheric responses during and after the storm. High‐latitude stations (KIL, Rothera) showed marked signal changes on 11 May, while the mid‐latitude site (ASB) exhibited delayed peaks on 13 May. The associated solar energetic particle (SEP) event produced noticeable VLF fluctuations about 10 min after onset, particularly evident in high‐resolution data. Wavelet analysis confirmed the presence of atmospheric gravity waves (15 s–30 min periods). SABER observations showed enhanced Nitric Oxide (NO) concentrations in the D‐ and E/F‐regions, peaking on 11 May at 85–95 km and 110–130 km altitudes, respectively. These NO increases, caused by energetic particle precipitation, were observed across both hemispheres. Temperature profiles revealed storm‐related warming above 95 km and cooling below 90 km in the Northern Hemisphere, linked to Joule heating, NO‐induced infrared cooling, and vertical transport. The study highlights the value of combining ground‐based VLF and satellite data to understand storm‐time ionospheric dynamics, emphasizing hemispheric asymmetries, delayed responses, and regional variability in space weather effects.

This study investigates the effects of geomagnetically induced currents (GIC) on New Zealand's electrical infrastructure via an intentional ground injection of direct current (DC) from a high voltage DC converter station. GIC manifests as quasi‐DC currents within power systems, potentially causing transformer saturation, increased reactive power demand, elevated harmonics, and transformer overloading. In January 2023, with support from Transpower, we utilized New Zealand's high voltage DC link to inject current directly into the ground at Haywards substation in Wellington, North island. This allowed us to monitor the effects of half‐cycle saturation of two 216 MVA, 220/110 kV autotransformers. Over 9 days, six injection tests were conducted, each lasting between one and 2 hours, with peak ground injection current reaching approximately 621 A, corresponding to 35 A measured to flow through one of the transformers on site. The saturation threshold of the transformer was measured to be around 20 A. Analysis of power quality measurements revealed a clear correlation between even total harmonic distortion and DC current, along with corresponding increases in acoustic sound and vibration levels. There was strong agreement between differential magnetometer (DMM) and transformer measurements of DC current. Variations in very low frequency radio harmonic observations were correlated with temporal variations in ground current. These findings provide valuable insights into the impact of GICs on in‐service equipment and inform grid operators on potential mitigation strategies.

Relativistic electrons are an important space weather hazard, being a major source of radiation damage to satellites and posing a risk to humans in space. We use approximately 20 years of data from the US Global Positioning System (GPS) satellite NS41 to determine the characteristics of the geomagnetic storms that lead to the largest relativistic electron fluxes in GPS orbit. The largest coronal mass ejection (CME)‐driven events are associated with the solar wind having negative excursions of the IMF with minimum values of −14 nT two hours prior to zero epoch, defined as the time of the minimum in the index, and strong minima, reaching −130 nT at zero epoch. In contrast, events driven by high speed solar wind streams (HSSs) are associated with smaller negative excursions of IMF with minimum values of −4 nT one to two hours prior to zero epoch and moderate minima, reaching −60 nT at zero epoch. Compared with HSS‐driven events, peak = 2.0 MeV fluxes associated with CME‐driven events are larger by factors of 1.3 at = 4.5 and 2.4 at = 6.5. Both the CME‐ and HSS‐driven events are associated with enhancements in the solar wind number density and pressure prior to zero epoch. Following zero epoch the solar wind number density and pressure become low and substorm activity is enhanced for several days.

Geomagnetically Induced Currents (GICs) in grounded, conducting infrastructure (e.g., power networks) represent an important space weather hazard. GICs are driven by the changing magnetic field at the Earth's surface. Due to a sparsity of GIC measurements, the rate of change of the horizontal geomagnetic field is often used as a proxy. We focus on one cause of large GICs: Sudden Commencements (SCs) caused by an increase in solar wind dynamic pressure. Despite appearing homogeneous in 1‐min cadence magnetic field data, a systematic variation has been observed in the correlation with the resulting GIC, depending on the magnetic local time. We investigate two questions: (a) do the data capture SC morphology, and (b) are some SCs intrinsically linked to larger GICs? We find that 1‐min magnetic field measurements underestimate the maximum , with systematic local time variation. We introduce an analytical model that describes the key components of an SC, the use of which strengthens the correlation between maximum and GIC during SCs . We conduct synthetic experiments with our analytical SC model and the GIC and geoelectric field models of Southern New Zealand and the United Kingdom, respectively. We find that the modeled GIC and geoelectric field linearly scale with the “size” of the SC, and non‐linearly with the “speed” of the magnetic change. Rotating the magnetic signature shifts the geoelectric field and GICs across the network. Forecasting the model parameters would enable robust forecasting of SC‐related GICs.

This study investigates the effects of geomagnetically induced currents (GIC) on New Zealand’s electrical infrastructure via an intentional ground injection of direct current (DC) from a high voltage DC converter station. GIC manifests as quasi-DC currents within power systems, potentially causing transformer saturation, increased reactive power demand, elevated harmonics, and transformer overloading. In January 2023, with support from Transpower, we utilised New Zealand’s high voltage DC link to inject current directly into the ground at Haywards substation in Wellington, North island. This allowed us to monitor the effects of half-cycle saturation on two 216 MVA, 220/110 kV autotransformers and the associated transmission lines. Over nine days, six injection tests were conducted, each lasting between one and two hours, with peak ground current reaching approximately 621 A, corresponding to 35 A measured to flow through one of the transformers on site. The saturation threshold was measured to be around 20 A. Analysis of power quality measurements revealed a clear correlation between even total harmonic distortion (THD) and DC current, along with corresponding increases in acoustic sound and vibration levels. There was strong agreement between differential magnetometer (DMM) and transformer measurements. Variations in very low frequency (VLF) radio harmonic observations were correlated with temporal variations in ground current. These findings provide valuable insights into the impact of GICs on in-service equipment and inform grid operators on potential mitigation strategies.

In early May 2024 6 Coronal Mass Ejections were launched toward the Earth in short succession. This triggered the G5 “Gannon” geomagnetic storm lasting from 17 UT on 10 May 2024 to 9 UT on 12 May 2024. Auroral displays were seen around the world including at lower latitudes than previous geomagnetic disturbances this century. Magnetic field variations measured at multiple sites in New Zealand exceeded values observed over the past 30 years. In 2022, following many years of industry‐research collaboration, an updated and enhanced “All of New Zealand” Geomagnetically Induced Current (GIC) mitigation strategy was developed. This strategy involves targeted line disconnections to reconfigure the network. The goal of this approach is to reduce GIC magnitudes and durations at the most at risk transformers, while still maintaining the continuous supply of power. Following training of control room staff, this strategy was declared operational in mid‐2023. Once disturbance levels reached the G5 threshold during the Gannon storm Transpower control room staff followed the planned procedure and implemented the mitigation strategy for the first time. There was no impact to New Zealand's electrical supply from this storm. GIC was measured at more than 70 transformers throughout the New Zealand power grid. Peak GICs observed after the mitigation strategy was in place reached 113 A at a transformer in the city of Dunedin. Without mitigation, modeling shows that a peak GIC of 200 A through the neutral would have occurred, exacerbated by the tripping of a transformer at a neighboring substation.

Geomagnetically induced current (GIC) measurements made at two 3 phase, 3 limb transformers, operating in the Halfway Bush substation in Dunedin, New Zealand have been analyzed during the May 2024 Gannon Storm. GIC measurements were combined with very low frequency radio wave AC harmonic measurements made nearby, and reactive power measurements made at key points in the substation. This study focuses on the 11 May, 00–14 UT period when geomagnetic activity was high and the 220 kV transformers, T6 and T3, experienced multiple short periods where GIC > 50 A in each transformer, maximizing at 113 A. During high GIC periods linear enhancements of even order AC harmonic intensity were identified, particularly for the 2nd and 4th harmonics, consistent with asymmetric half‐cycle transformer core saturation. Reactive power consumption (Q_con, MVAr) increased linearly when GIC levels were >30 A, consistent with the enhancement of even order AC harmonics due to transformer core saturation >30 A DC. Transformer T6 exhibited a reactive power response of 0.038 MVAr/A, while for T3 it was 0.026 MVAr/A. Simple extrapolation of these findings to extreme storm modeling of the New Zealand high voltage grid suggests that an additional ∼200–350 MVAr of generation would be required to compensate for peak increased reactive power consumption at 19 of the most affected sites during a Carrington‐level event. Such additional power requirements are likely to be within the capabilities of the power generation network.

The characteristics of very low frequency (VLF) radio wave propagation in the Earth‐ionosphere waveguide are determined particularly through dawn and dusk using phase and amplitude measurements of man‐made signals propagating below the ionospheric D region. For the first time variations of “Wait” height and sharpness parameters, H' and β, have been determined for dawn and dusk conditions. These measurements provide observational data to constrain D region modeling efforts, extending the capabilities of VLF propagation monitoring for geophysical phenomena such as lightning, solar flares, and energetic particle precipitation. At mid‐latitudes, H' varied from ∼85 km at night, then, starting from solar zenith angle (SZA) ∼ −97.5°, rapidly down to ∼73 km at dawn (SZA = −90°), then back up to ∼78 km at SZA ∼ −75° and then down to the appropriate noon value for the latitude (and season). In contrast, from noon through dusk to night, H' varied essentially monotonically from ∼70 to 75 km through ∼80 to ∼85 km. At low latitudes no dawn minimum in H' was observed, due to the reduced effect of galactic cosmic rays (GCR). Sharpness, β, varied from its nighttime value of ∼0.6 km⁻¹ down to a minimum of ∼0.25 km⁻¹ at SZA ∼85° near dusk or ∼75° near dawn, rising again to (SZA‐dependent) noon values of ∼0.35–0.5 km⁻¹. The results are interpreted through the geophysical effects controlling D region electrons, including the daytime dominant role of solar Lyman‐α from low to mid‐latitudes, and the greater role of GCR at increasingly higher mid‐latitudes.