Stochastic Environmental Research and Risk Assessment最新文献

This study assessed the risk of urban road waterlogging and the threshold of the influencing factors using software simulation and data analysis. This study selected the road space in the main urban area of Hangzhou City from 2019 to 2021 as the research object. ArcGIS software was used to study the spatial distribution of road waterlogging points. Kernel density analysis and the Geographic Detector (GD) method were used to determine the dominant factors affecting road waterlogging. This study reveals the central clustering distribution characteristics of road waterlogging and the five-level risk zoning of disasters. The simulation results show that the highest-risk areas for road waterlogging in the main urban area of Hangzhou are distributed in Chao Wang Road, Jianguo Middle Road, Jianguo South Road, Hupao Road, Lingyin Road, Fuchunjiang Road, Moganshan Road Sect. 4, and Tianmu Mountain Road Sect. 3. The ranking of the impact factors for road waterlogging was as follows: elevation > vegetation coverage > slope > impervious surface abundance > distance from rivers. Factor threshold for worst flooding is that the elevation of < 15–20 m, a slope of < 8–10°, vegetation coverage of < 10%, and an abundance of impermeable surfaces > 60–70%. Elevation and vegetation coverage were the significant factors with the greatest impact on road space waterlogging. The combination of elevation and vegetation coverage, elevation and slope, and elevation and impervious surface abundance had a greater impact on road waterlogging than the other three combinations. All the interactions of the influencing factors had a nonlinear enhancing effect on urban road waterlogging disasters.

In the field of groundwater, accurate delineation of contaminant plumes is critical for designing effective remediation strategies. Typically, this identification poses a challenge as it involves solving an inverse problem with limited concentration data available. To improve the understanding of contaminant behavior within aquifers, hydrogeophysics emerges as a powerful tool by enabling the combination of non-invasive geophysical techniques (i.e., electrical resistivity tomography—ERT) and hydrological variables. This paper investigates the potential of the Ensemble Smoother with Multiple Data Assimilation method to address the inverse problem at hand by simultaneously assimilating observed ERT data and scattered concentration values from monitoring wells. A novelty aspect is the integration of a Convolutional Neural Network (CNN) to replace and expedite the expensive geophysical forward model. The proposed approach is applied to a synthetic case study, simulating a tracer test in an unconfined aquifer. Five scenarios are compared, allowing to explore the effects of combining multiple data sources and their abundance. The outcomes highlight the efficacy of the proposed approach in estimating the spatial distribution of a concentration plume. Notably, the scenario integrating apparent resistivity with concentration values emerges as the most promising, as long as there are enough concentration data. This underlines the importance of adopting a comprehensive approach to tracer plume mapping by leveraging different types of information. Additionally, a comparison was conducted between the inverse procedure solved using the full geophysical forward model and the CNN model, showcasing comparable performance in terms of results, but with a significant acceleration in computational time.

Dust explosions occur frequently during production, transportation, and storage processes involving combustible dusts, with dust explosions caused by de-dusting systems being the most common. To prevent such accidents, we need to perform timely and accurate risk assessment. Therefore, we have developed a risk assessment model for dust explosion of dust duct deposition based on convolutional neural network with an attention mechanism (ConvNeXt-Tsc). By enhancing the ConvNeXt block and introducing an attention mechanism, we can more accurately extract the critical features related to the thickness of deposited dust in images of the ducts, achieving a model recognition accuracy of 95.15%. We have verified that the model has a high assessment accuracy in practical applications, which helps to detect potential hazards in dust ducts in time and avoid explosion accidents. The results show that the model has a wide range of application prospects in sedimentary dust explosion risk assessment, with high reliability, practicality, and scientific rigor.

In recent years, urban flooding has emerged as a major challenge, with land cover change identified as a key contributing factor. This study investigates the sensitivity of urban flooding risk to land cover changes. Seven urban land cover maps from different years and five different rainfall events, were selected as the examples. Based on hydrodynamic model simulations, this study analyzed the relationship between the total area of urban flooding and the proportion of ponding depths across various depth intervals and the land cover change. The study region was divided into 41 sub-areas based on road classifications and building clusters. The urban flood risk considering the aggregation of urban flooding, maximum ponding depth, the extent of the ponded area, and the average ponding depth was quantified within these sub-regions. Additionally, ten characteristic points were extracted from two sub-area with significant risk changes to analyze the logic of urban flooding risk evolution under land use change. The results indicate that: (1) There is a positive correlation between the total area of urban flooding and the proportion of high ponding depths and increasing impervious surfaces. (2) Urbanization significantly increases urban flooding risk, with 28 out of 41 areas experiencing heightened risk, including 6 sub-areas with risk increases exceeding 100%. (3) When the rainfall event changes from a 20-year to a 100-year return period, the maximum ponding depth in cropland stabilizes compared to impervious surfaces. Conversion of cropland to impervious surfaces accelerates increases in ponding depth and can lead to higher maximum ponding depths.

To improve slope reliability calculations and address high-nonlinearity in random fields, an AI algorithm, namely Convolutional Neural Network (CNN) with asymmetric convolution is introduced. The method accounts for the interdependence and auto-correlation of soil material and uses Python-based secondary development in ABAQUS Version 6.14 to improve computational efficiency and user-friendliness in finite element simulations. A Cholesky decomposition-based centroid point method is used for random fields to simplify computation. Additionally, an asymmetric convolution-based CNN surrogate model replaces finite element simulations to address challenges such as parameter correlations and random field discretization for improved analysis efficiency. The methodology uses random field samples and safety factors as inputs and outputs for training, which improves predictability and addressing high-dimensional issues. Its effectiveness is demonstrated through case studies involving single-layer undrained saturated clay slopes and double-layer cohesive soil slopes. The results demonstrate the effectiveness of the CNN approach that utilizes asymmetric convolution, with outcomes closely resembling those obtained through finite element simulation. This method demonstrates a 95.8% improvement in time efficiency compared to software-based calculations and a 93.5% enhancement over batch calculations using ABAQUS. These results confirm the effectiveness of the introduced reliability analysis method and the ability to provide accurate results while significantly boosting computational efficiency.

Making accurate and reliable predictions for monthly runoff in data scarce basins is still a major challenge. In this study, a new model, the CL-NDM, is developed by combining Convolutional Neural Network-Long Short-term Memory (CNN-LSTM) and a nonlinear dynamic model. The CL-NDM can overcome the deficiency of observed data by fusing spatial and temporal dependencies in runoff sequences at different stations. First, phase space reconstruction is used to enlarge the dimensions of the runoff sequences and reconstruct the attractors of the runoff sequences. Then, the CNN-LSTM is employed to construct the mapping between non-delay and delay attractors. Finally, the prediction set of the target variable is obtained by embedding multiple times. The CL-NDM is performed for monthly runoff prediction at eleven hydrological stations in the Weihe River, China. Compared with the CNN, LSTM and CNN-LSTM models, which require a large amount of training samples, the CL-NDM behaves much better, especially in situations with small training sample sizes. The maximum increase in R is 74%, and the maximum NSE is as large as 0.8. The maximum improvement in RMSE and MAPE is 53% and 88%, respectively. The CL-NDM has stronger ability to capture peak value while LSTM, CNN-LSTM and CNN models show obvious time lag in the prediction of peak point. The improved nonlinear dynamical model may provide a valuable method for runoff prediction in data-scarce regions.

This study proposes an integrated slope stability prediction model for various complex slope scenarios, including soil, rock, and rock-soil mixed situations. First, a small number of numerical slopes are constructed using the digital twin (DT) technique, and then these slope parameters are sorted and fine-tuned to build a database containing 19,666 soil, single/multiple sets of inclined joints, and rock-soil mixed slope scenarios. Second, the self-attention (SA) mechanism that can analyze the correlation of data features is connected to a classical convolutional neural network (CNN), forming a trained CNN-based SA model (CNN-SA) with 80% of the samples from the built database. The remaining 20% of the database and the stability of six actual slopes are then used for prediction. The performance of the CNN-SA is compared and evaluated. The results indicate that the DT technique is a reliable tool for providing the data to train the AI models, especially when the sample data is limited. As the complexity of the slopes increases, the prediction error of the models increases, and the CNN-based SA mechanism can effectively reduce these prediction errors compared to a classical CNN and other attention mechanisms.

In recent decades, the annual mean temperature has increased, with unusual alternations of hot and cold years. In addition, the changes in temporal precipitation patterns are caused by complex interactions between temperature change, the global water cycle, and other components of the Earth’s systems. To construct a statistical model of these temporal patterns in terms of temperature and precipitation, we propose a logistic mixture of spatial multivariate penalized regression splines for temporal profiles and apply this model to the contiguous United States climate data over 123 years (1900 to 2022) at 252 weather stations. The results reveal that the proposed model identifies climatologically meaningful clusters of weather stations in the contiguous United States with two important meteorological variables, temperature and precipitation, identifying the climate change patterns of each climate zone. The surface air temperature increased in the Northeast and West (Mountain and Pacific) regions, where the climate is affected by the continental Arctic air. A notable increment of precipitation also occurred in the Northeast. In contrast, the South region, where the climate is affected by the tropical Atlantic Ocean, is more stable than other regions in terms of year-to-year variations in temperature and precipitation.

Precise multi-step water level predictions are crucial for managing water resources and mitigating the effects of extreme weather. This study introduces a novel approach by integrating Variational Mode Decomposition (VMD), Whale Optimization Algorithm (WOA), and Long Short-Term Memory (LSTM) to forecast variations in water levels, employing both endogenous and exogenous environmental variables. Furthermore, this research proposes two additional fusion algorithms, each possessing unique potential for enhancement: Multivariate Long Short-Term Memory (MLSTM) and an advancement in the Residual Sequence (RESID). The predictive accuracy of these diverse algorithms is assessed using data from the water levels in Jinan Baotu Spring, China. The findings indicate that the VMD-WOA-LSTM model presents the most robust results for both long-term and short-term predictions. For multi-step, ultra-short-term forecasts, VMD-WOA-MLSTM proves to be a pragmatic algorithm. However, the refined algorithm that incorporates RESID does not significantly improve and, indeed, may diminish prediction accuracy. Conclusively, the VMD-WOA-LSTM, exemplifying a data-driven predictive algorithm, boasts high accuracy and demonstrates versatility in water level forecasting across various scenarios.

The accurate prediction of precipitation is crucial for agricultural management, water resources planning, and drought monitoring. One effective approach involves using a combination of linear and nonlinear models in a hybrid system. This study focuses on enhancing the hybrid model by employing the signal decomposition method, particularly for the complex nonlinear component. The research evaluated the effectiveness of incorporating seasonal autoregressive integrated moving average (SARIMA) with empirical mode decomposition (EMD) and maximal overlap discrete wavelet transform (MODWT) methods in the hybrid model structure using monthly precipitation data from stations in Iran. The procedure involved obtaining error series from the SARIMA model, decomposing the error series into intrinsic mode functions (IMFs) using EMD, and then applying support vector regression to forecast them. The evaluation criteria showed that using EMD in the hybrid model structure enhanced its efficiency by reducing significant error criteria and increasing residual predictive deviation. The proposed model also preserved precipitation forecasts in terms of time, with overestimated forecasts exhibiting high efficiency (RPD values > 2.5). Additionally, incorporating MODWT as a secondary decomposition in the final step of the proposed model further improved precipitation forecasting accuracy compared to the hybrid model solely incorporating EMD. The assimilation of signal decomposition methods in a hybrid model can enhance the accuracy and reliability of precipitation forecasts by revealing important error patterns.