•Professor Hiroyuki Nagahama,
specializing in Rock fracture mechanics, Earth continuum mechanics, and Earthquake prediction
•Professor Norihiro Nakamura,
specializing in Earth and planetary magnetism studies, Magnetic field, and Meteorites
•Associate Professor Jun Muto, specializing in Structural geology, and Experimental rock mechanics

Our research focuses on earthquakes and rock deformation.
Earthquakes and crustal deformation are complex deformation processes on Earth that occur in non-equilibrium systems where energy is not conserved. By integrating mathematical and physical theories, laboratory rock experiments, and field geological surveys, we aim to achieve a more comprehensive understanding of Earth's deformation processes beyond observational methods and scales.

How do earthquakes start and stop?
Electron microscope analysis of the Nojima Fault, which caused the 1995 Hyogo-ken Nanbu Earthquake, revealed that the fault plane became molten from frictional heating during the earthquake (Fig. 1). Although there are many active faults in Japan, not many of the faults that have caused earthquakes are exposed, so it is necessary to travel to other global sites to examine exposed faults. Figure 2 shows an outcrop of the Alpine Fault in the South Island of New Zealand. This outcrop presents a rare opportunity to see where the Pacific Plate (green rocks on the left) and the Australian Plate (brown rocks on the right) meet on land. The tremendous force of plate collision severely deforms the rocks on both sides.

Figure 1: Flow areas (top) and magnetic anomalies (bottom) developing in the Nojima fault. The red area, billow-like wavy folds, is a magnetic anomaly suggestive of frictional melting of the fault (modified from Fukuzawa et al., 2017).

Figure 2: Outcrop photograph of the NZ Alpine Fault. The Pacific Plate and the Australian Plate are in contact. With Associate Professor Tomoki Okada (Tohoku University, left) and Dr. Rick Sibson (Emeritus Professor, University of Otago, right).

We use a combination of various rock deformation apparatuses to investigate how rocks are deformed underground. Japan is deforming at a strain rate of 10-14/s owing to plate subduction. This is very slow deformation that brings two points 1 km apart only 0.3 mm closer in one year. Such slow deformation causes even hard rocks to flow slowly. Figure 3 shows a miniature fault with a magnitude of -4, which was created using an apparatus that can reproduce the flow of rocks deep underground. Adding water to this fault, we found that the strength of the fault was reduced to 1/10 to 1/100 of that without water.
On the other hand, it is known that when an earthquake occurs, rocks deform at a tremendous strain rate exceeding 1000/s. At such a high strain rate, rocks can be shattered into pieces in an instant (movie). By comparing the shape and particle size distribution of such pulverized samples with natural fault rocks, we investigate the strain rate and energy dissipation process during earthquakes.

Figure 3: Photograph of a miniature fault created in the experiment.

Movie: Rock crushing at high speed experiment

It is difficult to know precisely when and at what magnitude an earthquake will occur. Studies have shown, however, that significant earthquakes are accompanied by various anomalies, one of which is anomalies in concentrations of radioactive materials such as radon. Radon gas was found to have increased before the 1995 Kobe Earthquake, as measured at the radioisotope facility at Kobe Pharmaceutical University, located 20 km from the epicenter. Similar fluctuations also appeared before the 2003 Tokachi-Oki Earthquake and the 2011 Tohoku-Oki Earthquake. Using machine learning, we investigated whether these radon concentration fluctuations were statistically related to seismic activity, and found that radon concentration fluctuations are strongly related to the cumulative moment (energy) of earthquakes in a particular area. This revealed the possibility that the occurrence of microscopic cracks (fissures) in the bedrock before a major earthquake increases radon exhalation from the ground, increasing the radon concentration in the atmosphere (Fig. 4). Together with radioisotope facilities across the country, we are constructing a network for monitoring atmospheric radon and are taking on the challenge of data-driven prediction of seismic activity.

Figure 4: Image showing the exhalation of radon gas from underground. The concentration of radon in the atmosphere is measured at radio isotope facilities, where the radon containing gas dissipates to the ground through cracks (fissures) in the bedrock that occur before and after earthquakes.

One of the factors that make earthquakes and faults so difficult to predict is that Earth's crust is heterogeneous and contains many discontinuities. To address these challenges, we are working on modeling gravity and electromagnetic fields using continuum mechanics with a new gauge field similar to the Yang–Mills field. The study of electromagnetism associated with earthquakes has led to the study of magnetism in fault rocks, and then to other phenomena in planetary science, such as the generation of plasma magnetic fields, meteorite impacts, and the study of magnetic fields in the primordial solar nebula. Our research continues to develop in unexpected ways.