Explore the vast range of titles available.

Despite a century of research, turbulent flows mobilizing bedload remain elusive, while seismic waves generated by surface processes can unravel river dynamics. We studied the seismic signals emitted near rivers in two tributaries characterized by large boulders. Data show an unusually high dominant seismic frequency, reaching >2 times the frequency observed in nearby smoother channels. Consistent high‐frequency content during periods without bedload transport prompts the hypothesis that turbulence is a key contributor to generating higher frequencies. Assuming that dominant turbulent eddies decrease in size due to boulder‐constrained flow, we formulate a frequency scaling relationship that aligns well with field data. A positive relationship of the frequency with water depth breaks at bedload onset, indicating that dissipation of flow energy partitions between turbulence and bedload transport. Our study shows that seismic frequency captures contrasting bed morphologies in mountain streams, offering insights into flow‐roughness interactions. Plain Language Summary River processes, like water flow and sediment transport, generate energy that turns into seismic waves traveling through the ground. Studying these waves allows researchers to gain insights into how rivers function. We compared energy from rivers with large boulders to nearby streams with smoother surfaces. We found that boulder‐rich rivers produce higher seismic frequencies as the boulders reduce the size of turbulence‐related eddies. Our study shows how analyzing seismic energy helps to understand river dynamics, including flow and sediment transport. Key Points Boulder‐bed channels in the Liwu River, Taiwan, exhibit higher seismic frequency than channels with smoother beds The higher seismic frequencies are due to a reduction of turbulent eddy sizes constrained by boulder spacing The frequency‐depth relationship in seismic data shifts during and after bedload transport, likely due to energy partitioning or changes in bed roughness

With the global increase in demand for energy, energy conservation of research and development buildings has become of primary importance for building owners. Knowledge based on the patterns in energy consumption of previous years could be used to predict the near-future energy usage of buildings, to optimize and facilitate more effective energy consumption. Hence, this research aimed to develop a generic model for predicting energy consumption. Air-conditioning was used to exemplify the generic model for electricity consumption, as it is the process that often consumes the most energy in a public building. The purpose of this paper is to present this model and the related findings. After causative factors were determined, the methods of linear regression and various machine learning techniques—including the earlier machine learning techniques of support vector machine, random forest, and multilayer perceptron, and the later machine learning techniques of deep neural network, recurrent neural network, long short-term memory, and gated recurrent unit—were applied for prediction. Among them, the prediction of random forest resulted in an R2 of 88% ahead of the first month and 81% ahead of the third month. These experimental results demonstrate that the prediction model is reliable and significantly accurate. Building owners could further enrich the model for energy conservation and management.

Accurate forecasts of hourly water levels during typhoons are crucial to disaster emergency response. To mitigate flood damage, the development of a water-level forecasting model has played an essential role. We propose a model based on a dilated causal convolutional neural network (DCCNN) that can yield water-level forecasts with lead times of 1- to 6-h. A DCCNN model can efficiently exploit a broad-range history. Residual and skip connections are also applied throughout the network to enable training of deeper networks and to accelerate convergence. To demonstrate the superiority of the proposed forecasting technique, we applied it to a dataset of 16 typhoon events that occurred during the years 2012–2017 in the Yilan River basin in Taiwan. In order to examine the efficiency of the improved methodology, we also compared the proposed model with two existing models that were based on the multilayer perceptron (MLP) and the support vector machine (SVM). The results indicate that a DCCNN-based model is superior to both the SVM and MLP models, especially for modeling peak water levels. Much of the performance improvement of the proposed model is due to its ability to provide water-level forecasts with a long lead time. The proposed model is expected to be particularly useful in support of disaster warning systems.

A reservoir inflow forecasting system represents a crucial technique in reservoir operation and disaster prevention, particularly in areas where the primary water source derives from typhoon events. This includes the study area of the current research, i.e., the Shihmen Reservoir (Taiwan). Effectively depositing short and high-intensity rainfall and avoiding disaster losses present significant challenges in this regard. However, the high variability and uncertainty of such rainfall events make them difficult to forecast using traditional physical-based models, which require too many calculations for application in real-time disaster forecasting. Accordingly, in this study, seven machine learning (ML) algorithms, including three conventional ML and four deep learning algorithms, were compared to derive their effectiveness for reservoir inflow forecasting in extreme weather events. The forecasting lead-times were set to 1, 4, and 6-h, representing short, medium, and long-term forecasting, respectively. Moreover, to ensure the stability and credibility of the models, two types of integrated approaches, ensemble means and switched prediction method (SP) were also employed. The results showed that although an optimal algorithm could be selected for the short, medium, and long-term, individual algorithms did not always perform well in all events. Nonetheless, the integrated approaches can effectively combine the advantages of all the included algorithms and generate more accurate and stable forecasting results, particularly when using SP, which was involved in the top three performances among all typhoon examples and indicated the best average performance. In the short-term forecast, the RMSE of the testing events is 107.2 m3/s while using SP, ranking 3rd among all 9 methods. In the medium-term forecast, the RMSE predicted by the SP is 281.72 m3/s (Rank = 1). In the long-term forecast, the SP also performed the best among the 9 methods, and the RMSE was 477.14 m3/s. In conclusion, if only single model forecast is considered, gated recurrent unit, a type of transformed recurrent neural network, will yield the best performance. Furthermore, integrated forecasts, particularly involving SP, can effectively improve the accuracy and stability of forecasts to render a model more applicable to an actual situation.

Outburst flooding after a landslide dam breach causes global fatalities and devastation. Information on the timing, magnitude, and location of the landslide dam is crucial to hazard assessment. Despite recent efforts, successful real-time detection of landslide dams in mountain valleys and dam breakages is rare. Here, we present a series of seismic analysis including landslide detection, identification of landslide dam formations, and monitoring of dam breaches. We show the working of our analysis on a recent landslide dam that occurred in eastern Taiwan. The results indicate that our seismic analysis provides important information on the location and magnitude of landslides and the dam forming based on data acquired from a regional broadband seismic network. Furthermore, we see that the failure of the landslide dam is directly caught by the riverside seismic signals. To provide warning times for impending floods to downstream areas, we believe that proximal high-quality seismic signals along the river channel are viable options for an operational real-time monitoring system, for landslide dams occurring in mountain valleys. Our work can be a starting point to raise awareness in the community.

Seismic signals, with their remote and continuous monitoring advantages, have been instrumental in unveiling various landslide characteristics and have been widely applied in the past decades. However, a few studies have extended these results to provide geologists with pre-survey information, thus enhancing the understanding of the landslide process. In this research, we utilize the deep-seated Cilan landslide (CL) as a case study and employ a series of seismic analyses, including spectrogram analysis, single-force inversion, and geohazard location. These techniques enable us to determine the physical processes, sliding direction, mass amount estimation, and location of the deep-seated landslide. Through efficient discrete Fourier transforms for spectrograms, we identified three distinct events, with the first being the most substantial. Further analysis of spectrograms using a semi-log frequency axis generated by discrete Stockwell transform revealed that Event 1 consisted of four sliding failures occurring within 30 s with decreasing sliding mass. Subsequent Events 2 and 3 were minor toppling and rockfalls, respectively. Geohazard location further constrained the source location, indicating that Events 1 and 2 likely originated from the same slope. Subsequently, the sliding direction retrieved from single-force inversion and the volume estimation were determined to be 153.67° and 557 118 m3, respectively, for the CL. Geological survey data with drone analysis corroborated the above seismological findings, with the sliding direction and source volume estimated to be around 148° and 664 926 m3, respectively, closely aligning with the seismic results. Furthermore, the detailed dynamic process observed in the spectrogram of Event 1 suggested a possible failure mechanism of CL involving advancing, retrogressing, enlarging, or widening. By combining the above mechanism with geomorphological features identified during field surveys, such as the imbrication-like feature in the deposits and the gravitational slope deformation, with video from the event, we can infer the failure mechanism of retrogression of Event 1 after shear-off from the toe. Then, the widening activity was caused by the failure process for subsequent events, like Events 2 and 3. This case study underscores the significance of remote and adjacent seismic stations in offering seismological-based landslide characteristics and a time vision of the physical processes of landslides, thereby assisting geologists in landslide observation and deciphering landslide evolution.

Accurate real-time forecasts of inundation depth and extent during typhoon flooding are crucial to disaster emergency response. To manage disaster risk, the development of a flood inundation forecasting model has been recognized as essential. In this paper, a forecasting model by integrating a hydrodynamic model, k-means clustering algorithm and support vector machines (SVM) is proposed. The task of this study is divided into four parts. First, the SOBEK model is used in simulating inundation hydrodynamics. Second, the k-means clustering algorithm classifies flood inundation data and identifies the dominant clusters of flood gauging stations. Third, SVM yields water level forecasts with 1–3 h lead time. Finally, a spatial expansion module produces flood inundation maps, based on forecasted information from flood gauging stations and consideration of flood causative factors. To demonstrate the effectiveness of the proposed forecasting model, we present an application to the Yilan River basin, Taiwan. The forecasting results indicate that the simulated water level forecasts from the point forecasting module are in good agreement with the observed data, and the proposed model yields the accurate flood inundation maps for 1–3 h lead time. These results indicate that the proposed model accurately forecasts not only flood inundation depth but also inundation extent. This flood inundation forecasting model is expected to be useful in providing early flood warning information for disaster emergency response.

Bedload transport is a natural process that strongly affects the Earth's surface system. An important component of quantifying bedload transport flux and establishing early warning systems is the identification of the onset of bedload motion. Bedload transport can be monitored with high temporal resolution using passive acoustic methods, for example, hydrophones. Yet, an efficient method for identifying the onset of bedload transport from long‐term continuous acoustic data is still lacking. Benford's Law defines a probability distribution of the first‐digit of data sets and has been used to identify anomalies. Here, we apply Benford's law to continuous acoustic recordings from Baiyang hydrometric station, a tributary of Liwu River, Taroko National Park, Taiwan at the frequency of 32 kHz from stationary hydrophones deployed for 3 years since 2019. We construct a workflow to parse sound combinations of bedload transportation and analyze them in the context of hydrometric sensing constraining the onset, and recession of bedload transport. We identified three separate sound classes in the data related to the noise produced by the motion of pebbles, water flow, and air. We identify two bedload transport events that lasted 17 and 45 hr, respectively, covering about 0.35% of the total recorded time. The workflow could be transferred to other different catchments, events, or data sets. Due to the influence of instrument and background noise on the regularity of the residuals of the first‐digit, we recommend identifying the first‐digit distribution of the background noise and ruling it out before implementing this workflow. Plain Language Summary Long‐term, high‐frequency monitoring of Earth surface processes brings huge data sets that can be of high quality. Benford's Law defines the specific probability distribution of the first‐digit of the data sets and has been used to identify anomalies and high‐energy events. We provide a workflow for applying Benford's Law to identify the onset of the motion of coarse sediment along the river bed at a time resolution of seconds. Since Benford's Law has demonstrated usefulness in acoustic amplitude analysis in this study, it could serve as a tool for identifying anomalous events in any kind of real‐time data series, which could be beneficial for generating event samples for machine learning applications. Key Points Long‐term, high‐frequency acoustic monitoring constitutes huge‐volume data sets with a low signal‐to‐noise ratio The distinct first‐digit distribution between signal and noise can used to filter out 99% of background noise from acoustic recordings We applied the method to a three‐year‐long acoustic data set in Baiyang, identifying two bedload transportation events

Hazards from gravity-driven instabilities on hillslope (termed ‘landquake’ in this study) are an important problem facing us today. Rapid detection of landquake events is crucial for hazard mitigation and emergency response. Based on the real-time broadband data in Taiwan, we have developed a near real-time landquake monitoring system, which is a fully automatic process based on waveform inversion that yields source information (e.g., location and mechanism) and identifies the landquake source by examining waveform fitness for different types of source mechanisms. This system has been successfully tested offline using seismic records during the passage of the 2009 Typhoon Morakot in Taiwan and has been in online operation during the typhoon season in 2015. In practice, certain levels of station coverage (station gap < 180°), signal-to-noise ratio (SNR ≥ 5.0), and a threshold of event size (volume >10 6  m 3 and area > 0.20 km 2 ) are required to ensure good performance (fitness > 0.6 for successful source identification) of the system, which can be readily implemented in other places in the world with real-time seismic networks and high landquake activities.

Global climate is thought to be modulated by the supply of minerals to Earth’s surface. Whereas silicate weathering removes carbon dioxide (CO 2 ) from the atmosphere, weathering of accessory carbonate and sulfide minerals is a geologically relevant source of CO 2 . Although these weathering pathways commonly operate side by side, we lack quantitative constraints on their co-variation across erosion rate gradients. Here we use stream-water chemistry across an erosion rate gradient of three orders of magnitude in shales and sandstones of southern Taiwan, and find that sulfide and carbonate weathering rates rise with increasing erosion, while silicate weathering rates remain steady. As a result, on timescales shorter than marine sulfide compensation (approximately 10 6 –10 7 years), weathering in rapidly eroding terrain leads to net CO 2 emission rates that are about twice as fast as CO 2 sequestration rates in slow-eroding terrain. We propose that these weathering reactions are linked and that sulfuric acid generated from sulfide oxidation boosts carbonate solubility, whereas silicate weathering kinetics remain unaffected, possibly due to efficient buffering of the pH. We expect that these patterns are broadly applicable to many Cenozoic mountain ranges that expose marine metasediments. Unlike sulfide and carbonate, silicate weathering does not increase with physical erosion, which could result in a net release of carbon dioxide associated with uplift, according to stream-water chemistry of southern Taiwan.