Abstract Impacts of temperature assimilation on the Venus atmosphere are investigated by the observing system simulation experiments assuming cross-link radio occultation measurements of the Venus atmosphere (CROVA). Synthetic temperature data are prepared by a Venusian general circulation model (VGCM) with 80% of the solar heating to produce a weaker super-rotation than the default model would. Observational points of radio occultation measurements are realistically determined by a pair of satellites with feasible orbits. Then the vertical temperature profiles in 40–90 km altitudes at those points are assimilated into the original VGCM where the faster super-rotation is reproduced. The results show that the three-dimensional structure of the true atmosphere could be significantly improved not only in temperature but also in horizontal winds, when temperature observations are available with more than 40 points per day using 3 small cube satellites for 30 days. It would be promising to plan CROVA mission for future.

Abstract Equatorial sea surface temperature anomalies have profound impacts on tropical and extra-tropical sub-seasonal and seasonal weather variability. While extensive research explored the impacts of El Niño-Southern Oscillation (ENSO) and its diversity on the global teleconnections, the weakness of ENSO-attributed signals over Europe, Africa and the Middle East has remained a challenge. In this study, ENSO teleconnection modulation is defined by considering the concurrent state of the equatorial Atlantic and Indian Oceans to investigate their impacts on modulating tropical and extra-tropical teleconnections. The results show that ENSO teleconnection patterns are influenced by the anomalous state of the two other basins, in some cases even leading to out-of-phase circulation anomalies. The impacts are highlighted for circulation, temperature, and precipitation anomalies in all seasons, aiming to understand the modulation of ENSO teleconnections by other equatorial oceanic basins, ultimately facilitating sub-seasonal and seasonal weather forecast skill.

Abstract East Asian monsoon shows strong annual cycle with the wet season in summer and dry in winter. In addition to annual cycle, some monsoon regions exhibit distinct climatological sub-seasonal variations such as fast annual cycle, climatological intraseasonal oscillations (CISOs) and climate singularities. In this paper a significant cyclic-type outgoing longwave radiation (OLR) singularity during the spring to summer transition period over the South China Sea (SCS) is identified based on the daily OLR data from 1979 to 2023. The quasi-CISO in each individual year is invented as an analogue to CISO. Its relationship with Madden–Julian Oscillation (MJO) and the SCS summer monsoon onset (SCSSM) is analyzed. The quasi-CISO shows large interannual variability. A group of years with the quasi-CISO amplitudes larger than the medium of all sample years are selected to form the quasi-CISO mode. Among them 72% of the quasi-CISO cyclic valley coincides with the SCSSM onset time represented by the persistent intensification of the monsoonal westerlies, which are the SCS-CISO years. The MJO activity during the sixteen SCS-CISO years shows three different levels. Composite results of eight SCS-CISO years with strong MJO show clear synchronized planetary- and regional-scale convection and circulation progressions where strong MJOs over the equatorial Indian Ocean appeared two weeks before SCSSM onset. Strong convection over the SCS associated with the SCSSM onset process can be triggered by the enhanced low-level southerly flow by MJO’s subsidence leg over the equatorial maritime continent through increasing the moisture transport. This study provides observational-based evidence that supports earlier findings about CISO, CMJO and SCSSM onset. The findings can be applied to evaluate the Sub-seasonal to Seasonal forecast model products for better understanding of the extended-range weather and climate predictability.

Abstract Downstream sweep erosion (DSE), rapid erosion at gorge entrances where wide channels narrow into confined gorges, drives rapid bedrock gorge removal. However, sediment can both enhance erosion through abrasion and decelerate it by channel aggradation, yet the role of sediment dynamics in modulating DSE efficiency remains poorly understood. We used SfM-MVS photogrammetry with aerial photographs and UAV imagery from 2009 to 2025 to quantify upstream gorge retreat and sediment-driven feedbacks in Taiwan’s Daan River Gorge. DSE-driven gorge retreat rate decreased dramatically from 11.5 m/yr (2009–2014) to 0.7 m/yr (2014–2016) due to channel aggradation, before partially recovering to 1.5 m/yr (2016–2025) following alluvial bar scouring. High sediment flux initially enhanced erosion through the tool effect but later inhibited it by forming bars that deflected flow and shifted impact angles from perpendicular to parallel. This cyclic mechanism extended gorge lifespan from ~ 50 years to 182–208 years, though removal remains faster than the ~ 470-year tectonic recurrence. These findings demonstrate that sediment can strongly regulate gorge evolution and transient topography persistence in the Daan River Gorge.

Abstract Based on 393 cases, this study developed an enhanced light gradient boosting machine (KOA–LightGBM) model integrated with the Kepler optimization algorithm (KOA), employing SHAP method to improve interpretability. Key factors including slope height, angle, unit weight, cohesion, internal friction angle, and pore water pressure ratio were determined to evaluate the slope stability under the combined effect of geometric, geotechnical, and hydraulic conditions. Five metrics systematically compared pre-/post-optimization performance of SVM, DT, CatBoost, and LightGBM models, indicating KOA substantially improved the classification capability of all models, with optimized model accuracy not lower than 0.80. Among them, the KOA–LightGBM model performs best, with the accuracy and AUC values of 0.91 and 0.97, respectively, a logarithmic loss of only 0.28, a runtime not exceeding 7.5 s, and run-to-run variations within 0.1 s. The influence mechanism of each feature on the model output is revealed by the SHAP method, which identifies cohesion, internal friction angle, and unit weight as having positive correlations with the prediction results, whereas slope height, slope angle, and pore water pressure ratio are found to be negatively correlated. Among them, cohesion has the most significant effect on the prediction results of the KOA–LightGBM model, followed by internal friction angle, slope angle, slope height, unit weight and pore water pressure ratio.

Abstract As an island with high tectonic uplift and a humid climate with frequent typhoons, landsliding is one of the primary processes that shapes the landscape in Taiwan. To investigate the impact of landslides on long-term erosion in a landslide-dominated setting, we analyzed the spatial and temporal variations in the erosion rate in the Lao-Nung basin, southwestern Taiwan. In this basin, a series of well-surveyed and dated fluvial and landslide fan terraces are distributed along its main stem, providing an ideal setting to reconstruct paleo-erosion rates using in situ 10 Be concentrations in quartz. The results reveal relatively steady long-term erosion rates averaging ~ 0.7 mm/yr but with notable temporal and spatial variations. A reduced erosion rate around ~ 12 ka likely reflects cold and dry climatic conditions, as supported by pollen and organic matter records in lake sediments. In contrast, an elevated rate around ~ 6 ka corresponds to a warm and wet climate and increased landslide activity, evidenced by the widespread formation of landslide fan terraces. However, the absence of such terraces in the upstream sampling area suggests the higher erosion rate may also reflect spatial differences in uplift, as indicated by increasing normalized channel steepness index upstream. Despite extensive recent landsliding, modern erosion rates remain relatively uniform along the main stem except in the upstream sections, suggesting that much of the landslide-derived sediments remain trapped in the upstream area rather than being efficiently transported downstream.

Abstract This study analyzes the effects of interannual variation in surface heat flux on the variability of the upper layer circulation in the East Sea (Sea of Japan) via a series of numerical experiments. Comparing the intrinsic variability of the upper layer circulation, the interannual variation in the surface heat flux amplifies the variability in the Yamato Basin, but not in the Ulleung Basin. The variability in the water temperature in the northern region is highly correlated with the variability in the surface heat flux with a 1-month time lag. Under strong cooling conditions, the cyclonic gyre and convection in the northern region are intensified, and the expansion of the cold water area facilitates a more pronounced southward intrusion compared to weak cooling conditions. However, the interannual variation in surface heat flux does not significantly impact the northernmost latitude of the western boundary current (i.e., the East Korea Warm Current). Instead, the northernmost latitude of the East Korea Warm Current is influenced primarily by the magnitude of the strong positive winter wind stress curl in the northern region, which greatly affects the southward flow of cold water along the Korean coast. An increased (decreased) volume transport or lateral heat flux through Korea/Tsushima Strait promotes the eastward (westward) propagation of eddies distributed in the Yamato Basin, thereby affecting the meandering pattern of the Tsushima Warm Current. This study is expected to enhance understanding of the respective contributions of interannual variation in individual external forcing factors to the variability of the upper layer circulation in the East Sea.

The magnitude of major earthquakes induced by active faults is essential to seismic hazard analysis. However, the sample size of such earthquakes recorded by the accurate, modern instrumentation is minimal. Hence, the estimate using the (conventional) Frequentist method only based on samples/observations becomes less specific. Like other applications of the Bayesian approach that integrates (limited) observation and prior information, this paper introduces a new one for estimating the magnitude of major earthquakes induced by an active fault. In addition to the Bayesian algorithm, two case studies were also presented in this paper.

We report a rarely observed case of steady aseismic deformation in the context of a fold-and-thrust belt, with a well-documented structural and lithological background. We focus on a 12-km-long section across the foothills of southwestern Taiwan, where about 23 mm/yr of westward compression is observed. From west to east, the surface geological structures include an anticline, a thrust and a backthrust. We determine Holocene uplift rates based on fluvial terraces, invert the interseismic 3D velocity field using existing geodetic datasets, and build a geological cross-section to constrain the possible deep geometry for the structure responsible for the observed surface deformation. Geodetic vertical velocities and Holocene uplift rates show a similar pattern, with rates rapidly increasing eastward, then remaining relatively constant across the fold axis and thrust, and finally sharply decreasing across the backthrust, across which InSAR (Interferometric Synthetic Aperture Radar) observations suggest a velocity discontinuity. Our observations show that active deformation is occurring mainly aseismically and involves the anticline and backthrust. Our cross-section illustrates a 4–5 km deep wedge with a passive roof thrust corresponding to the backthrust, on the hanging wall of which the anticline is located. A classical fault-bend fold model with a slip rate of 21 ± 2 mm/yr can explain most of the observations, yet local misfit suggests a possible contribution to uplift from pure shear of clayey rocks in the anticline core. Based on published records from a deep well drilled across the fold core and backthrust, clay-rich lithology and elevated fluid content are likely to favor aseismic slip.