Existing landslide susceptibility prediction methods often fail to fully account for the spatial heterogeneity of environmental factors such as topography, soil, and vegetation, nor do they accurately reflect the impact of extreme rainfall on landslide susceptibility. To overcome these limitations, this study proposes two innovative methods. First, to address the issue of spatial heterogeneity, a Deep Embedding Clustering (DEC) approach is introduced. DEC utilizes an autoencoder to map environmental factors to a lower-dimensional space, capturing nonlinear relationships between variables and performing clustering in this space. Unlike traditional methods, DEC does not rely on simple distance measures; instead, it jointly optimizes clustering centers and feature representations, enabling more precise regional delineation, which significantly enhances prediction accuracy and adaptability to varying environments. Second, to address the static nature of rainfall thresholds, a mixed distribution modeling strategy is proposed for both non-extreme and extreme rainfall. In this strategy, non-extreme rainfall is modeled using the Gamma distribution to describe cumulative effects, while extreme rainfall is modeled using the Generalized Pareto Distribution (GPD) to model extreme values, with thresholds dynamically determined using the Pickands theorem. Additionally, a Bayesian online parameter updating mechanism is implemented to dynamically adjust distribution parameters, recalibrating the model when real-time rainfall data deviates from historical distributions, significantly reducing response time and improving the model's adaptability to changing rainfall patterns. By combining Deep Embedding Clustering (DEC) and the mixed distribution rainfall threshold model, this study achieves more precise spatial zoning and dynamic rainfall responses, greatly improving prediction accuracy and timeliness. Compared to traditional models relying on uniform thresholds, the experimental results show that landslide density and event numbers have increased from 0.038 events/km2 and 44 events to 0.044 events/km2 and 59 events, respectively, validating the importance of incorporating spatial heterogeneity and distinct rainfall event types in landslide susceptibility prediction.
This study proposes a method for determining earthquake focal depths by combining the sPn phase with the waveform cross-correlation technique, based on waveform data recorded by the Fujian Seismic Network from the 2024 M 7.3 Hualien offshore earthquake and the 2025 M 6.2 Tainan earthquake. The Pn phase onset was precisely aligned using waveform cross-correlation, and the arrival time difference (Delta t) between the sPn and Pn phases was extracted via a sliding time-window correlation method. The focal depths were derived using a layered velocity model for the Taiwan region. Results show that the calculated focal depth for the Hualien earthquake is 23.1 km (Delta t = 6.9 s), with a relative error of 2.7% compared to the official result (22.5 km) from the Central Weather Administration of Taiwan. For the Tainan earthquake, the depth is 17.9 km (Delta t = 6.1 s), with a relative error of 13.3%. In this study, we show that a cross-correlation threshold of 0.8 and a bandpass filtering of 0.1-0.3 Hz are efficient to suppress noise and significantly improve depth accuracy for shallow earthquakes with depth <30 km. Compared to traditional travel-time location methods, this approach exhibits superior noise resistance and computational efficiency. Future work will focus on optimizing 3D velocity structures, integrating multiple phases, and applying deep learning techniques such as convolutional neural networks, aiming to improve the results in a more reliable and automatic way, and to provide efficient support on earthquake emergency response.
Ghana and its surrounding West African region are often regarded as a stable intraplate area far from active plate boundaries, characterized by relatively low seismicity compared to tectonically active zones like the East African Rift. However, the region has experienced significant earthquakes (including a magnitude ∼6.5 event in 1939 in Ghana) and a growing number of small-to-moderate events in recent decades, underscoring a critical need to resolve its seismotectonic framework. Seismic hazard studies have largely overlooked West Africa, and the causes and tectonic controls of its earthquakes remain poorly understood. To address this, we synthesize geological, geophysical, and seismic data in a comprehensive review to examine the regional tectonic setting, the spatial distribution of historical and recent seismicity, stress field characteristics, and the status of earthquake monitoring. We conclude that West African intraplate earthquakes are predominantly driven by the reactivation of pre-existing crustal weaknesses under far-field plate forces, augmented by stress transfer from the Atlantic margin (along offshore transform fracture zones) and the Cameroon Volcanic Line. Our observations highlight the region’s current geodynamics and provide a basis for improved seismic hazard assessment. Given the rapidly growing population and infrastructure in seismically vulnerable areas across West Africa, these findings are crucial for informing earthquake preparedness and risk mitigation strategies.
Recently, with rapid developments in distributed acoustic sensing (DAS) techniques, optical fibers are widely used as seismometers to carry out dense and broadband seismic observation. However, the frequency responses, especially the low-frequency responses of submarine cables, which are of key importance in focal mechanism and deep seismic imaging, are still poorly understood. In this study, we comprehensively evaluated the phase and amplitude responses of the submarine DAS array by comparing teleseismic records from the DAS array and traditional broadband seismometers. Our 21 km long DAS array is located around 1 500 km from the epicentre, and two fixed seismometers from the Shandong Provincial Seismic Network are positioned at either end of the array. Our results show that for phase response, the data from DAS and seismometers are of great similarity, with a cross-correlation coefficient reaching 80 %. For amplitudes, the velocity value recorded by CHD (Shandong Provincial Seismic Network Changdao Seismic Station) seismometers is about 1.6 times the value obtained by DAS. In terms of surface wave dispersion, the differences between the different data sets are generally less than 3%. Our results verify the sensitivity of submarine optical fibers to low-frequency seismic signals and the feasibility of deep structure imaging using submarine DAS arrays.
Loess is a Quaternary aeolian sediment that has been deposited over an undulating bedrock landscape, forming the Chinese Loess Plateau. Where this bedrock comprises mudstones, the loess-bedrock interface conditions the landscape in a state of high landslide susceptibility. In this neo-tectonically active region, earthquakes and rainfall are highly likely to act as triggers. This study focuses on a representative loess–mudstone interface landslide located on a hillside behind the Vehicle Management Office in Wuqi County, Yan’an City, Central China, to investigate the coupled effects of rainfall infiltration and seismic loading on slope deformation and failure mechanisms. Numerical simulations were developed to better understand how co-seismic landslides are affected by different rainfall conditions, and how this resulted in different mechanisms of failure and runout. The results show that rainfall infiltration progressively weakens the mudstone layer, leading to the upward extension of the potential slip zone and a marked reduction in slope stability. Under subsequent seismic loading, deformation is concentrated along the loess–mudstone contact, and topographic amplification further increases peak ground acceleration at the slope crest. The interaction between rainfall-induced softening and seismic excitation significantly accelerates slope failure, highlighting the critical control of the mudstone layer’s post-softening strength on overall stability. The findings of this study provide a valuable reference for analyzing the disaster mechanisms of similar loess landslides.
