Stochastic Environmental Research and Risk Assessment最新文献

SPAtial EFficiency (SPAEF) metric is one of the most thoroughly used metrics in hydrologic community. In this study, our aim is to improve SPAEF by replacing the histogram match component with other statistical indices, i.e. kurtosis and earth mover’s distance, or by adding a fourth or fifth component such as kurtosis and skewness. The existing spatial metrics i.e. SPAtial efficiency (SPAEF), structural similarity (SSIM) and spatial pattern efficiency metric (SPEM) were compared with newly proposed metrics to assess their converging performance. The mesoscale hydrologic model (mHM) of the Moselle River is used to simulate streamflow (Q) and actual evapotranspiration (AET). The two-source energy balance AET during the growing season is used as monthly reference maps to calculate the spatial performance of the model. The moderate resolution imaging spectroradiometer based leaf area index is utilized by the mHM via pedo-transfer functions and multi-scale parameter regionalization approach to scale the potential ET. In addition to the real monthly AET maps, we also tested these metrics using a synthetic true AET map simulated with a known parameter set for a randomly selected day. The results demonstrate that the newly developed four-component metric i.e. SPAtial Hybrid 4 (SPAH4) slightly outperforms conventional three-component metric i.e. SPAEF (3% better). However, SPAH4 significantly outperforms the other existing metrics i.e. 40% better than SSIM and 50% better than SPEM. We believe that other fields such as remote sensing, change detection, function space optimization and image processing can also benefit from SPAH4.

Ecosystem services (ESs) refer to the benefits that humans obtain from ecosystems. These services are subject to environmental changes and human interventions, which introduce a significant level of uncertainty. Traditional ES modeling approaches often employ Bayesian networks, but they fall short in capturing spatiotemporal dynamic change processes. To address this limitation, dynamic Bayesian networks (DBNs) have emerged as stochastic models capable of incorporating uncertainty and capturing dynamic changes. Consequently, DBNs have found increasing application in ES modeling. However, the structure and parameter learning of DBNs present complexities within the field of ES modeling. To mitigate the reliance on expert knowledge, this study proposes an algorithm for structure and parameter learning, integrating the InVEST (Integrated Valuation of Ecosystem Services and Trade-Offs) model with DBNs to develop a comprehensive understanding of the spatiotemporal dynamics and uncertainty of ESs in the Tarim River Basin, China from 2000 to 2020. The study further evaluates the tradeoffs and synergies among four key ecosystem services: water yield, habitat quality, sediment delivery ratio, and carbon storage and sequestration. The findings show that (1) the proposed structure learning and parameter learning algorithm for DBNs, including the hill-climb algorithm, linear analysis, the Markov blanket, and the EM algorithm, effectively address subjective factors that can influence model learning when dealing with uncertainty; (2) significant spatial heterogeneity is observed in the supply of ESs within the Tarim River Basin, with notable changes in habitat quality, water yield, and sediment delivery ratios occurring between 2000–2005, 2010–2015, and 2015–2020, respectively; (3) tradeoffs exist between water yield and habitat quality, as well as between soil conservation and carbon sequestration, while synergies are found among habitat quality, soil retention, and carbon sequestration. The land-use type emerges as the most influential factor affecting the tradeoffs and synergies of ESs. This study serves to validate the capacity of DBNs in addressing spatiotemporal dynamic changes and establishes an improved research methodology for ES modeling that considers uncertainty.

Boundary conditions play a crucial role in groundwater contamination source identification (GCSI), but they may be complex and reliable estimates are difficult to obtain in advance in actual situations. If the estimated values deviate significantly from the actual situation, the GCSI results will be inaccurate. However, very few studies have attempted to identify the boundary conditions in GCSI, and even when they are identified, they are often considered too simple. The boundary mode (B_mode) is assumed to be known, but in reality, it is often unknown and is more complex than initially assumed. Previous practices based on this assumption may not accurately reflect actual situations. Therefore, this study focused on the concentration boundaries, and the boundary conditions were also considered unknown variables, along with contamination source information and model parameters. To alleviate the problem of identifying the boundary conditions under an unknown B_mode, we proposed for the first time to treat the B_mode as an unknown variable. Thus, the source information, model parameters, B_mode, and corresponding parameters in the boundary concentration (BC) function were identified simultaneously. The Differential Evolution Adaptive Metropolis with a Snooker Update and Sampling from a Past Archive (DREAM_(ZS)) algorithm and a Kriging surrogate model were used as the primary means of solution. We designed four different synthetic cases to test the effectiveness of the above ideas. When identifying the B_mode, the obtained BC mostly fitted well with the true BC. It was therefore considered feasible for identifying the B_mode. The performance of the DREAM_(ZS) algorithm was found to be superior to the traditional DREAM algorithm and was more efficient.

Forecasting accurate water quality is very important in aquaculture, environment monitoring, and many other applications. Many internal and external factors influence water quality. Therefore, water quality parameters exhibit complex time series characteristics. Consequently, long-range accurate prediction of water quality parameters suffers from poor propagation of information from past timepoints to further future timepoints. Moreover, to synchronise the prediction model with the changes in the time series characteristics, periodic retraining of the prediction model is required and such retraining is to be done on resource-restricted computation devices. In this work, we present a low-cost training approach to improve long-range multi-step water quality prediction. We train a short-range predictor to save training effort. Then, we strive to achieve and/or improve long-range prediction using multi-step iterative ensembling during inference. Experimental results on 9 water quality datasets demonstrate that the proposed method achieves significantly lower error than the existing state-of-the-art approaches. Our approach significantly outperforms the existing approaches in several standard metrics, even in the case of future timepoints at long distances.

Precise prediction of runoff is not only conducive to the prevention of floods and droughts but also to the rational use of water resources. Due to the frequency of weather extremes and the complexity of runoff variability, achieving accurate runoff predictions is challenging. This research develops a deep-learning ensemble model for interval prediction based on meteorological and hydrological factors. The model can be divided into four stages: feature extraction, decomposition, point prediction, and interval prediction. First, Pearson's correlation coefficient filters out key driving variables affecting runoff. Next, the original data are decomposed by variational modal decomposition (VMD) to intrinsic modal function (IMF); Then, each IMF is decomposed by complementary ensemble empirical modal decomposition (CEEMD) to capture more data details. Following, the runoff point prediction portion is realized by the attention mechanism fusion gated recurrent unit (AM-GRU). In this study, data from Dunhuang and Panjiazhuang stations, located in the upper and lower reaches of the Shule River in China, were used to validate and analyze the VMD-CEEMD-ISSA-AM-GRU (VCIAG) model. The results show that (1) the VCIAG model has the best fitting effect which the NSE values of Dunhuang and Panjiazhuang stations are 0.97 and 0.96, respectively. (2) In the multi-period prediction in advance, the highest prediction accuracy is achieved when the prediction period is 1 day and the accuracy of the prediction decreases gradually as the prediction period becomes longer. (3) In flood early warning, the VCIAG performs well at both stations, which suggests that the proposed model can take precautionary measures in advance before the floods come. (4) In terms of interval prediction, the VCIAG model has the narrowest prediction interval width and the highest prediction accuracy, which enhances the application value of the model.

An urban water demand with high accuracy and reliability plays fundamental role in creating a predictive water supply system. This is necessary to implement mechanisms and systems that can be used to estimate water demands. The paper aims to forecast the amounts of clean water given to metropolis with single multiplicative neuron model artificial neural network. For this purpose, a regular data set of Istanbul metropolis was selected as a model and utilised in the application process. Single multiplicative neuron model artificial neural network is a neural network that does not have the hidden layer unit number problem that many shallow and deep artificial neural networks have. In this study, the black hole optimization algorithm is used for the first time in the literature for the training of the single multiplicative neuron model artificial neural network. In line with the analysis results, it is concluded that the proposed new approach achieved better prediction results than the other compared methods for the time series of amounts of clean water given to metropolis.

This study presents an analysis of the influence of climatic conditions and the operation of a dam reservoir on the occurrence of ice cover and water temperature in two rivers (natural and transformed by reservoir operations) located in the Carpathian Mountains (central Europe). The analyses are based on data obtained from four hydrological and two climatological stations. The Extreme Gradient Boosting (XGBoost) machine learning model was used to quantitatively separate the effects of climate change from the effects arising from the operation of the dam reservoir. An analysis of the effects of reservoir operation on the phase synchronization between air and river water temperatures based on a continuous wavelet transform was also conducted. The analyses showed that there has been an increase in the average air temperature of the study area in November by 1.2 °C per decade (over the period 1984–2016), accompanied by an increase in winter water temperature of 0.3 °C per decade over the same period. As water and air temperatures associated with the river not influenced by the reservoir increased, there was a simultaneous reduction in the duration of ice cover, reaching nine days per decade. The river influenced by the dam reservoir showed a 1.05 °C increase in winter water temperature from the period 1994–2007 to the period 1981–1994, for which the operation of the reservoir was 65% responsible and climatic conditions were 35% responsible. As a result of the reservoir operation, the synchronization of air and water temperatures was disrupted. Increasing water temperatures resulted in a reduction in the average annual number of days with ice cover (by 27.3 days), for which the operation of the dam reservoir was 77.5% responsible, while climatic conditions were 22.5% responsible.

Statistical descriptions of extreme met-ocean conditions are essential for the safe and reliable design and operation of structures in marine environments. The significant wave height (({H}_{S})) is one of the most essential wave parameters for coastal and offshore structural design. Recent studies have reported that a time-varying component exists globally in the ({H}_{S}). Therefore, the non-stationary behavior of an annual maximum series of ({H}_{S}) is important for various ocean engineering applications. This study aims to analyze the frequency of ({H}_{S}) over the northern Indian Ocean by modeling the non-stationarity in the ({H}_{S}) series using a non-stationary Generalized Extreme Value (GEV) distribution. The hourly maximum ({H}_{S}) data (with a spatial resolution of 0.5° longitude × 0.5° latitude) collected from the global atmospheric reanalysis dataset of the European Centre for Medium-Range Weather Forecasts (ECMWF) is used for the study. To model the annual maximum series of ({H}_{S}) using a non-stationary GEV distribution, two physical covariates (El-Ni (widetilde{n}) o Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD)) and time covariates are introduced into the location and scale parameters of the GEV distribution. The return levels of various frequencies of ({H}_{S}) are estimated under non-stationary conditions. From the results, average increases of 13.46%, 13.66%, 13.85%, and 14.02% are observed over the study area for the 25-year, 50-year, 100-year, and 200-year return periods, respectively. A maximum percentage decrease of 33.3% and a percentage increase of 167% are observed in the return levels of various return periods. The changes in the non-stationary return levels over time highlight the importance of modeling the non-stationarity in ({H}_{S}).

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.