Mount Bora in Central Sulawesi, Indonesia, represents a promising, underexplored geothermal field located within a tectonically active zone. This study introduces a low-cost, non-invasive approach using Horizontal-to-Vertical Time–Frequency Analysis (HVTFA) for integrated geothermal characterization and seismic site-response assessment. A total of 25 microtremor measurements were collected across the study area, of which 24 produced HVTFA-derived ellipticity curves that satisfied the adopted clarity and reliability criteria. These curves were used to extract key site-characterization parameters, including the fundamental frequency (f0), the HVTFA-derived relative index (Kg), calculated from the ellipticity peak amplitude and fundamental frequency, and the average shear-wave velocity in the upper 30 m (Vs30). HVTFA was applied to address spectral instability commonly encountered in classical HVSR analysis, particularly in geologically complex environments. Rayleigh wave ellipticity inversion was then used to derive 1D Vs profiles and construct interpolated 2D cross-sections. The results reveal pronounced lateral heterogeneity in subsurface conditions, with zones of relatively low Vs and locally elevated Kg values concentrated in the southern sector near geothermal manifestations. These areas show relatively low Vs and locally elevated Kg values that may reflect spatial variations in subsurface conditions associated with geothermal processes. In contrast, higher Vs and f0 values in the northern and eastern parts of the study area indicate more competent subsurface materials with relatively stiffer subsurface conditions. The integrated HVTFA framework provides a low-cost framework for characterizing shallow subsurface heterogeneity and identifying zones that may be associated with geothermal influence, and supports microzonation efforts in volcanically and tectonically dynamic environments.
Landslide risk assessment is a core research field for landslide risk management and loss mitigation, and recent studies have increasingly employed artificial intelligence (AI)-based models. However, such models attribute variable importance only after model fitting, offering limited support for deciding which environmental variables should be prioritized for mitigation. Accordingly, this study applied Partial Least Squares Structural Equation Modeling (PLS-SEM), which specifies variable relationships a priori and estimates them within an explicit structural framework. For the Chungju–Jecheon region of Korea, eight candidate latent variables were defined—terrain morphology, terrain position, terrain ruggedness, hydrological concentration, hydrological flow, precipitation, soil characteristics, and forest characteristics—and a PLS-SEM model was constructed to analyze the structural relationships between the Landslide Susceptibility Index (LSI) and these variables. Model evaluation identified two final models, in both of which terrain morphology and soil characteristics were consistently the two dominant influences on the LSI, though their order varied. Model 1, comprising terrain morphology, soil characteristics, and forest characteristics, yielded R² = 0.444, whereas Model 2, in which forest characteristics were replaced by precipitation, showed slightly lower explanatory power (R² = 0.394). Explanatory power remained close to, but below, the moderate threshold, and categorical factors such as geological structure and land cover, together with the dynamic characteristics of precipitation, were not sufficiently captured. Future work should integrate static susceptibility with dynamic triggering, couple statistical and physically based models, and perform spatiotemporal and transfer validation across watersheds, thereby enhancing the reliability and practical applicability of the model.
This study evaluates the provenance and relative contributions of terrigenous sediments during the late Pleistocene–Holocene using clay mineral compositions from the northern Ulleung Basin, East Sea. The East Sea is highly sensitive to sea-level fluctuations, which alter ocean circulation and modulate the Chinese loess influx driven by the East Asian winter monsoon. From the glacial period (> 14.5 ka) to the Holocene (0–11.5 ka), smectite content increased from an average of 26.9 wt
Geological disposal is widely regarded as a viable long-term management strategy for high-level radioactive waste. Because repository performance is assessed over timescales extending from tens of thousands to, in some cases, up to one million years, potential environmental change is an important consideration in repository safety assessment. Future climatic, geological, hydrological, and ecological processes may influence groundwater flow systems, radionuclide transport pathways, biosphere conditions, and the performance of natural and engineered barriers. Consequently, national repository programs have developed various approaches to incorporate long-term environmental evolution into safety assessments. This review examines how long-term environmental change is addressed in geological disposal programs in Sweden, Finland, Switzerland, France, Canada, and Japan. The review compares national approaches to representing future environmental conditions and discusses common challenges associated with environmental evolution, biosphere and surface-system representation, integration of geosphere, biosphere, and engineered systems, and uncertainty over geological timescales. Although the reviewed programs differ in geological setting, repository concept, climatic history, and regulatory framework, several common themes emerge. Long-term environmental assessments employ scenario-based representations of future conditions, consider interactions among environmental and repository-system processes, and apply structured approaches for addressing uncertainty. At the same time, the environmental processes emphasized in safety assessments vary according to site-specific conditions and programmatic priorities. The reviewed cases suggest that approaches to long-term environmental assessment are shaped by both site-specific considerations and broader assessment objectives. While no single assessment framework is suitable for all repository settings, experience from existing national programs provides useful perspectives for the development of future repository programs and long-term environmental assessment frameworks, including those for Korea.
Rainfall-induced landslides pose significant geohazard risks, yet their occurrence can be anticipated through systematic instrumentation and historical data analysis. A thorough understanding of geological and geotechnical properties facilitates the identification of critical thresholds, which is essential for the development of landslide early warning systems (LEWS). This study presents experimental findings from a laboratory-simulated rainfall-induced landslide using a custom-designed tilting flume. The investigation focuses on variations in gravimetric moisture content (GMC), excess pore water pressure (EPWP), surface particle motion analyzed via particle image velocimetry (PIV), and the loss of fine content under controlled conditions. Experimental trials were conducted at three soil densities, with rainfall intensities of 150 mm/h and 300 mm/h. Instrumentation placed at the slope face and toe recorded significant variations across all monitored parameters, indicating their suitability for LEWS deployment. In contrast, minimal variations were observed at the heel portion. The analysis establishes a relationship between the relative factor of safety (FOS) degradation and variations in GMC, based on the principles of unsaturated soil mechanics. The results identify key parametric thresholds for landslide prediction and early warning in silty sand slopes. They show that instability in suction-dominated silty sand is governed not by rainfall intensity alone, but by its combined interaction with dry density, suction loss, localized moisture accumulation, EPWP development, and kinematic deformation.
