Search

In site-specific site-response assessments, observation-based site-specific approaches requiring a target–reference recording pair or a regional recording network cannot be implemented at many sites of interest. Thus, various estimation techniques have to be used. How effective are these techniques in predicting site-specific site responses (average over many earthquakes)? To address this question, we conduct a systematic comparison using a large data set which consists of detailed site metadata and Fourier outcrop linear site responses based on observations at 1725 K-NET and KiK-net sites. We first develop machine learning (i.e. random forest ( RF)) amplification models on a training data set (1580 sites). Then we test and compare their predictive powers at 145 independent testing sites with that of the one-dimensional (1D) ground response analysis (GRA). The standard deviation of residuals between observations and predictions, that is, between-site (site-to-site or inter-site) variability, is used as the benchmark. Results show that the machine learning model using a few predictor variables, surface roughness, peak frequency fP, HV, VS30, and depth Z2.5 achieves better performance than the physics-based modeling (GRA) using detailed 1D velocity profiles. This implies that machine learning can be more effective in using existing site information than 1D GRA which is inflicted by a high level of parametric and modeling uncertainties. This finding warrants the further exploration of machine learning in site effect characterization, especially on model transferability across different regions.

The computation of time series for hard-rock site conditions, either as input motion for site response calculations or for applications to installations constructed on this site type, is a crucial step for seismic hazard assessment. The current state-of-the-art is to apply physics-based corrections to surface motions to eliminate the influence of site effects for retrieving the underlying bedrock motion. Here we evaluate the application of the generalized inversion method (GIT) for deconvolving surface recordings to obtain hard-rock time series at the amplification-free seismic bedrock. The method includes an event-specific phase scaling approach on the surface recordings that takes changes in the signal duration between the surface and the seismic bedrock into account.We choose a total of 90 KiK-net surface-downhole sites for validation which do not have a significant velocity contrast below the downhole sensors and with the latter being located at sufficient depth so that they are not significantly impaired by downgoing waves. By comparing the empirical predictions from surface recordings with time series that have been recorded at the downhole sensors, we assess the accuracy of the predictions. In contrast to empirical and one-dimensional modelling approaches, which significantly overestimate the level of hard-rock ground motion for frequencies greater than a few Hz, we find quite high correlations and small variations in both spectral shape and amplitude over the entire frequency range. This approach delivers a data set that facilitates the development of reference ground-motion models., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

The northern and central part of Japan facing the Japan Sea are widely snow-covered through winter. There are two types of snowfall distribution: one is the plain-type snowfall dominant in the Japan-Sea coastal plain areas and the other is the mountain-type snowfall dominant along mountain ranges. Although snowfalls are not so heavy in the coastal areas, rare extreme snowfalls are caused by the Japan Sea polar air-mass convergence zone (JPCZ) developed over the southern part of Japan Sea, where the JPCZ tends to appear connecting dents of cyclonic isobars. Further, we found that those extreme snowfalls are often also associated with anticyclonic isobars appeared along the Japan-Sea coast of Japan. Those cyclonic and anticyclonic isobars form “stomach shape pattern over the Japan Sea. In addition, a polar low (PL) tends to appear off the Japan-Sea coast in the northern Japan. Ascending flows with near surface convergence with the JPCZ and PL over the Japan Sea are in conjunction with divergence in the mid troposphere, which causes mid tropospheric convergence in between. This process requires descending flow and causes near surface divergence over the central part of the Japan Sea, which resultantly increases in surface pressure and forms an anticyclonic circulation along the Japan-Sea coast in the northern Japan. Thus convergent snow-cloud band is formed along the coast extending to southwestward, and it encounters the JPCZ oriented to eastward, where extreme snowfall events occur. Further numerical simulations also support robust features., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

This is an open-source site database for a total number of 1742 earthquake recording sites in the K-NET (Kyoshin network) and KiK-net (Kiban Kyoshin network) networks in Japan. This database contains site characterization parameters directly derived from available velocity profiles, including average wave velocities, bedrock depths, and velocity contrast. Meanwhile, it also consists of parameters obtained from earthquake horizontal-to-vertical spectral ratio (HVSR), e.g., peak frequency, amplitude, width and prominence. In addition, the site database also comprises topographic and geological proxies inferred from regional models or maps. Each parameter is derived in a consistent manner for all sites. This site database can benefit the application of machine learning techniques in studies on site amplifications. Besides, it can facilitate, for instances, the search of the optimal site parameter(s) in depicting site amplification, the development and testing of new or existing prediction models or methodologies. This zip file contains earthquake HVSR data, including metadata and plots of automatically identified peaks (frequency, amplitude, prominence and width) at 1700+ K-NET and KiK-net strong-motion recordings stations. This MATLAB code is for automatically picking peaks (frequency, amplitude, prominence and width) from HVSR curves.

OBJECTIVE Diabetes mellitus is becoming a major cause of premature disability in Japan, and peripheral neuropathy is a common complication of diabetes. The aim of this study was to evaluate the relationship between the results of nerve conduction studies (NCS) and the size of the nerve determined by sonography in diabetic patients. METHODS Twenty diabetic patients (mean age +/- SD, 57.1 +/- 13.6 years) and 20 healthy volunteers (mean, 61.1 +/- 8.9 years) were enrolled in this study. Patients' wrists that had symptoms of carpal tunnel syndrome were not included in the study; those that were included had negative Phalen test results. We then divided the patients into 2 groups (patients with and without diabetic symmetric polyneuropathy [DPN]). The cross-sectional area (CSA) was measured in the carpal tunnel 5 cm proximal to the wrist and elbow joint of the median nerve. RESULTS There was a significant increase in the CSA in patients with DPN in the carpal tunnel compared with the control participants (P < .01) and patients without DPN (P < .01). The CSA in the carpal tunnel showed a significant correlation with the motor nerve conduction velocity (r = -0.473). CONCLUSIONS The CSA of the median nerve in the carpal tunnel of patients with DPN is greater than that in patients without DPN and healthy individuals and correlates with NCS.