We present a deep‐learning approach for earthquake detection using waveforms from a seismic array consisting of multiple seismographs. Although automated, deep‐learning earthquake detection techniques have recently been developed at the single‐station level, they have potential difficulty in reducing false detections owing to the presence of local noise inherent to each station. Here, we propose a deep‐learning‐based approach to efficiently analyze the waveforms observed by a seismic array, whereby we employ convolutional neural networks in conjunction with graph partitioning to group the waveforms from seismic stations within the array. We then apply the proposed method to waveform data recorded by a dense, local seismic array in the regional seismograph network around the Tokyo metropolitan area, Japan. Our method detects more than 97% of the local seismicity catalog, with less than 4% false positive rate, based on an optimal threshold value of the output earthquake probability of 0.61. A comparison with conventional deep‐learning‐based detectors demonstrates that our method yields fewer false detections for a given true earthquake detection rate. Furthermore, the current method exhibits the robustness to poor‐quality data and/or data that are missing at several stations within the array. Numerical experiments using subsampled data demonstrate that the present method has the potential to detect earthquakes even when half of the normally available seismic data are missing. We apply the proposed method to analyze 1‐h‐long continuous waveforms and identify new seismic events with extremely low signal‐to‐noise ratios that are not listed in existing catalogs. We also show the potential portability of the proposed method by applying it to seismic array data not used for the training.

We are constructing a tandem type electron cyclotron resonance (ECR) ion source (ECRIS). High-energy electrons in ECRIS plasma affect electron energy distribution and generate multicharged ion. In this study, we measure electron energy distribution function (EEDF) of low energy region (≦100 eV) in ECRIS plasma at extremely low pressures (10−3–10−5 Pa) by using cylindrical Langmuir probe. From the result, it is found that the EEDF correlates with the electron density and the temperature from the conventional probe analysis. In addition, we confirm that the tail of EEDF spreads to high energy region as the pressure rises and that there are electrons with high energy in ECR multicharged ion source plasma. The effective temperature estimated from the experimentally obtained EEDF is larger than the electron temperature obtained from the conventional method.

The discovery of slow slip events (SSEs) based on the installation of dense geodetic observation networks has provided important clues to understanding the process of stress release and accumulation in subduction zones. Because SSEs with durations of days to weeks do not often result in sufficient displacements that can be visually inspected, refined automated detection methods are required to understand the occurrence of SSEs. In this study, we propose a new method based on which SSEs can be detected in observations derived by a Global Navigation Satellite System (GNSS) array by using l₁ trend filtering, a variation of sparse estimation, in conjunction with combined ‐value techniques. The sparse estimation technique and data‐driven determination of hyperparameters are utilized in the proposed method to identify candidates of the event onsets. In addition, combined ‐value techniques are used to provide confidence values for the detections. The synthetic test demonstrated that the new method can detect 22 events of total 23 events and has only 1% false detections with the detection threshold 68%. The proposed method was then applied to daily displacements obtained at 39 GNSS stations in the Nankai subduction zone in western Shikoku, Southwest Japan. The results revealed that, in addition to all known events, new 12 events can be detected with the proposed method. Finally, we found the number of low‐frequency earthquakes in the target region increased around at the onsets of potential events.

For predictive evaluation based on quasi-posterior distributions, we develop a new information criterion, the posterior covariance information criterion (PCIC). PCIC generalizes the widely applicable information criterion (WAIC) so as to effectively handle predictive scenarios where likelihoods for the estimation and the evaluation of the model may be different. A typical example of such scenarios is the weighted likelihood inference, including prediction under covariate shift and counterfactual prediction. The proposed criterion uses a posterior covariance form and is computed by using only one Markov chain Monte Carlo run. Through numerical examples, we demonstrate how PCIC can apply in practice. Further, we show that PCIC is asymptotically unbiased to the quasi-Bayesian generalization error under mild conditions in weighted inference with both regular and singular statistical models.

We introduce a novel class of strictly proper scoring rules for statistical models on spheres that does not require the calculation of normalizing constants. The class contains the Hyvärinen scoring rule investigated by Mardia et al.(2016).