Broadband magnetotelluric (MT) measurements were conducted in 2010 and 2011 in the vicinity of Shinmoe‐dake Volcano in the Kirishima volcano group, Japan, where sub‐Plinian eruptions took place 3 times during 26–27 January 2011. By combining the new observations with previous MT data, it is found that an anomalous phase in excess of 90° is commonly observed in the northern sector of the Kirishima volcano group. Because the anomalous phase is not explained by 1‐D or 2‐D structure with isotropic resistivity media, 3‐D inversions were performed. By applying small errors to the anomalous phase, we successfully estimated a 3‐D resistivity structure that explains not only the normal data but also the anomalous phase data. The final model shows a vertical conductor that is located between a deep‐seated conductive body (at a depth greater than 10 km) and a shallow conductive layer. By applying the findings of geophysical and petrological studies of the 2011 sub‐Plinian eruptions, we infer that the subvertical conductor represents a zone of hydrothermal aqueous fluids at temperatures over 400°C, in which a magma pathway (interconnected melt) is partially and occasionally formed before magmatic eruptions. To the north of the deep conductor, earthquake swarms occurred from 1968 to 1969, suggesting that these earthquakes were caused by volcanic fluids.

Asama Volcano is an andesitic composite volcano and one of the most active volcanoes in Japan. In order to reveal electrical resistivity structure beneath the volcano accurately, we performed a 3-D inversion of dense magnetotelluric survey data. In order to prevent misinterpretation of the subsurface resistivity due to the steep topography around Asama Volcano, we used an unstructured tetrahedral mesh to represent the topography. Furthermore, we reduced the calculation time by transforming the inverse problem from the model space into the data space. Comparison of the new data-space method to the original model-space method showed that the calculation time required to update the model parameters was reduced as a result of the transformation, whereas the resistivity structure obtained remained unchanged. In the subsurface resistivity structure around Asama Volcano that was estimated from the inversion, resistive bodies were discovered to be located under the old eruption centres. In particular, under the 24 ka collapse caldera to the west of the presently active crater, a spherical resistive body was found to exist in isolation. In addition, there was a widespread conductive layer below the resistive surface layer. By comparison with previous hydrological and geochemical studies, the conductive layer was interpreted as being a high-water-content layer and an overlying layer rich in altered clay minerals. Because the western part of the volcanic conduit was considered to be the resistive area, which is inferred to consist of unfractured rocks with lower permeability than their surroundings, it would appear that the area obstructs the westward flow of the hydrothermal fluid beneath the summit, thereby contributing to higher concentrations of SO₄²⁻ and Cl⁻ in the spring water at the northern and eastern feet as well as the uneven location of a diffuse CO₂ anomaly.

We recorded volcanic lightning generated by Vulcanian explosions at Sakurajima Volcano using a synchronized multiparametric array. Physical properties of lightning are related to plume dynamics, and associated electromagnetic field variations are revealed by video observations (high speed and normal speed) together with infrasound and high sampling rate magnetotelluric signals. Data show that volcanic lightning at Sakurajima mainly occurs in the plume gas thrust region at a few hundred meters above the crater rim, where the overpressure of the turbulent volcanic jets determines the electrification of particles generating a complex charge structure in the growing plume. Organization of charges may be achieved at later stages when the plume transitions from the jet phase to the convective phase. Comparison with atmospheric sounding and maximum plume height data show that the effect of hydrometeors on flash generation at Sakurajima is negligible and can be more prudently considered as an additional factor contributing to the electrification of volcanic plumes.

The lightning generated by explosive volcanic eruptions is of interest not only as a promising technique for monitoring volcanic activity, but also for its broader implications and possible role in the origin of life on Earth, and its impact on the atmosphere and biosphere of the planet. However, at present the genetic mechanisms and physical properties of volcanic lightning remain poorly understood, as compared to our understanding of thundercloud lightning. Here, we present joint magnetotelluric (MT) data and video imagery that were used to investigate the physical properties of electrical discharges generated during explosive activity at Sakurajima volcano, Japan, and we compare these data with the characteristics of thundercloud lightning. Using two weeks of high-sensitivity, high-sample-rate MT data recorded in 2013, we detected weak electromagnetic signals radiated by volcanic lightning close to the crater. By carefully inspecting all MT waveforms that synchronized with visible flashes, and comparing with high-speed (3000 frame/s) and normal-speed (30 frame/s) videos, we identified two types of discharges. The first type consists of impulses (Type A) and is interpreted as cloud-to-ground (CG) lightning. The second type is characterized by weak electromagnetic variations with multiple peaks (Type B), and is interpreted as intra-cloud (IC) lightning. In addition, we observed a hybrid MT event wherein a continuous weak current accompanied Type A discharge. The observed features of volcanic lightning are similar to thunderstorm lightning, and the physical characteristics show that volcanic lightning can be treated as a miniature version of thunderstorm lightning in many respects. The overall duration, length, inter-stroke interval, peak current, and charge transfer all exhibit values 1–2 orders of magnitude smaller than those of thunderstorm lightning, thus suggesting a scaling relation between volcanic and thunderstorm lightning parameters that is independent of the type of charged particles. On the other hand, the polarities, which are estimated by long-time (3.4 yrs) MT (32 samples/s) and video (30 frame/s) observations, are different than those of normal thunderstorm lightning. These observations are consistent with the notion that charge structures in volcanic ash plumes are highly disordered and are characterized by numerous small charged regions with high charge density