Stochastic Environmental Research and Risk Assessment最新文献

Upper and lower percentiles of Flow Duration Curves (FDCs) of daily streamflow data were investigated to develop frequency curves. Upper percentiles with exceedance probability of 1, 5 and 10% (Q₁, Q₅, Q₁₀) were used for high flows, and lower percentiles with non-exceedance probability of 10, 5 and 1% (Q₉₀, Q₉₅, Q₉₉) for low flows. Median value (Q₅₀) was covered to represent the average conditions of streamflow. A mixed frequency analysis based on the total probability theorem taking zero values into account was applied for the lower percentiles of FDC. Case studies were performed for three intermittent Streamflow Gauging Stations (SGSs) from Kucuk Menderes River Basin in western Turkey. An overall assessment of results shows that the best-fit probability distribution function does not change from one SGS to another considerably for low flows while each SGS has its own probability distribution function for high flows. Upper and lower percentiles, and median value were calculated at various return periods by using the identified probability distribution functions. The calculated values were plotted in the form of frequency curves of high flow percentiles and low flow percentiles. The frequency curves have a practically significant potential use in hydrological analysis, water resources management and hydraulic design under high and low flow conditions. They are yet open to further development for regionalization and their applicability can be extended to ungauged sites in river basins.

Reliable estimation of reference evapotranspiration (ET_o), an essential component of optimal irrigation management, is challenging in many regions due to its complex dependence on meteorological factors. Alternative empirical models, often used to estimate ET_o considering data limitations, provide highly unreliable estimates for Iraq. This study aimed to formulate simpler empirical models for accurate ET_o estimation with fewer variables in different climate regions of Iraq. The metaheuristic Whale Optimization Algorithm (WOA) was used to finetune the coefficients of the nonlinear least square fitting regression (NLLSF) model during development. Two simpler models were developed based on (1) only mean air temperature (T) (NLLSF-T) and (2) solar radiation and T (NLLSF-R) as inputs. The performance of the models was validated using historical ground observations (2012–2021), and the ET_o was estimated using the Penman–Monteith method from the reanalyzed (ERA5) datasets (1959–2021). The models' spatial, seasonal, and temporal performance in estimating daily ET_o was rigorously evaluated using multiple statistical metrics and visual presentations. The Kling-Gupta Efficiency (KGE) and normalized root mean square error (NRMSE) of the NLLSF-T model were 0.95 and 0.30, respectively, compared to 0.75 and 0.40 for Kharrufa, the best-performing temperature-based models in Iraq. Similarly, NLLSF-R improved the KGE from 0.78 to 0.97 in KGE and NRMSE from 0.44 to 0.22 compared to Caprio, the best-performing radiation-based model in Iraq. The spatial assessment revealed both the models' excellent performance over most of Iraq, except in the far north, indicating their suitability in estimating ET_o in arid and semi-arid regions.

The intensities and frequencies of extreme rainstorms are increasing, which may result in severe inundation of urban metro systems. Although there is some risk assessment research on regional metro systems based on spatiotemporal data, the characteristics of specific metro stations and shortcomings in the emergency response process need more consideration. In this paper, a risk analysis model for rainstorm inundation in metro systems based on a Bayesian network and a practical case study are proposed. First, the risk factors are obtained by integrating general mechanism analysis and the case study. Second, an event evolution diagram is established to represent the comprehensive evolution process of a potential event. Third, the risk analysis model is established using a Bayesian network model considering the quantitative causal relationships between risk factors. This model is used to analyze the risk of supporting emergency management, including emergency preparation based on critical risk factor sensitivity identification, prewarning response strategy development based on risk analysis as rainstorms occur, and rescue strategy development based on risk analysis as rainstorm water flows into metro tunnels. Furthermore, this model can be flexibly improved as natural hazards and metro systems change and as new problems are exposed in practical cases.

This article provides a precise, quantitative description of the sampling error on burn counts in Monte-Carlo wildfire simulations - that is, the prediction variability introduced by the fact that the set of simulated fires is random and finite. We show that the marginal burn counts are (very nearly) Poisson-distributed in typical settings and infer through Bayesian updating that Gamma distributions are suitable summaries of the remaining uncertainty. In particular, the coefficient of variation of the burn count is equal to the inverse square root of its expected value, and this expected value is proportional to the number of simulated fires multiplied by the asymptotic burn probability. From these results, we derive practical guidelines for choosing the number of simulated fires and estimating the sampling error. Notably, the required number of simulated years is expressed as a power law. Such findings promise to relieve fire modelers of resource-consuming iterative experiments for sizing simulations and assessing their convergence: statistical theory provides better answers, faster.

The duration characteristics of near-fault earthquake ground motions play a significant role in the dynamic response of a structure. Linear regression-based models are extensively used to forecast ground motions and duration parameters. However, such an approach fails to account for the complexity arising from the non-linear patterns in the data set. Nevertheless, implementing machine learning algorithms has the ability to uncover these unexplored patterns as well as the unique characteristics of ground motions comprised in the datasets. In this study, statistical relationships between several duration metrics and intensity measures of near-fault ground motions are evaluated using machine learning algorithms. Four different machine learning algorithms, namely Regression, Decision Tree, Support Vector machines, and Gaussian Process regression model are trained to determine the optimum model. All these machine learning models were examined using the selected database of 200 near-fault pulse-like ground motions, which was split into two parts, with 75% of data used for training and the remaining 25% for testing. The results indicate that the fine tree model for bracketed duration, stepwise linear regression model for uniform duration, and the exponential and rational gaussian process regression model for significant and effective duration, showed more accurate and reliable results as compared to other models.

In developing countries, evaluating irrigation water quality using conventional methods can be costly and time-consuming. To overcome these challenges, this study explores the potential of utilizing physical parameters and artificial intelligence (AI) models for predicting and evaluating the quality indicators of irrigation water in aquifer systems. To achieve this goal, novel hybrid methods, namely the Whale Optimization Algorithm (WOA) and Wind-Driven Optimization (WDO), are employed in conjunction with Artificial Neural Network (ANN) models. The specific objective of this study is to forecast the Sodium Adsorption Ratio (SAR) by considering independent variables such as Na⁺, Mg²⁺, Ca²⁺, Na percent, K⁺, SO₄²⁻, Cl⁻, pH, and HCO₃⁻. A dataset of 540 samples from the Shiraz plain, collected over a statistical period of 16 years (2002–2018), is used to estimate the groundwater quality variables. A pre-processing technique is applied in the AI approach to enhance the model's efficiency. The results indicate that the WDO-ANN model exhibits higher accuracy (R² = 0.9983 and RMSE = 0.10618) than the WOA-ANN model (R² = 0.9957 and RMSE = 0.16957). The optimization of computational parameters and comparison of AI model structures demonstrate that the WDO-ANN model outperforms the WOA-ANN model in predictive ability. In general, using AI models as a tool for low-cost and timely prediction of underground water quality using physical parameters as input variables has a high potential.

Count time series exhibiting seasonal patterns are frequently encountered in practical scenarios. For example, the number of hospital emergency service arrivals may show seasonal behavior (Moriña et al. 2011 Stat Med 30:3125–3136). Numerous models have been proposed for the analysis of seasonal count time series with an unbounded support, yet seasonal patterns in bounded count time series, which are sometimes suffered in environmental science such as the number of monthly rainy-days and air quality level data, have not received formal attention. The contribution of this article lies in coping with the modeling challenges associated with seasonal count time series with a bounded support, which is beneficial for enhancing the applicability of environmental science data. This is achieved by introducing a seasonal structure and seasonally varying model parameters into the first-order binomial autoregressive (BAR(1)) model (McKenzie 1985 J Am Water Resour Assoc 21:645–650). The probabilistic and statistical properties, marginal distribution and some special cases of the proposed model are studied. Estimation of model parameters is conducted using the Yule-Walker, conditional least squares and maximum likelihood methods. The asymptotic normality of the estimators is also presented. To demonstrate the utility of our model in environmental data, applications are carried out on the monthly number of rainy-days in two Russian cities.

Accurate streamflow prediction is significant for water resources management. However, due to the impact of climate change and human activities, accurately identifying the input factors of the streamflow prediction model and achieving high-precision results presents a significant challenge. In this study, past streamflow, meteorological, and climate factors were utilized as inputs to develop a predictive scenario for the bi-decomposition of input factors and streamflow series, i.e. Scenario 3 (S3). Mutual information (MI) was applied to recognize the input factors prediction potential. Based on the predictive potentials, factors were progressively incorporated into the kernel extreme learning machine (KELM) and hybrid kernel extreme learning machine (HKELM) models optimized by the gazelle optimization algorithm (GOA) to ascertain the optimal input configuration for each sub-series. The prediction results of S3-KELM and S3-HKELM models were obtained by reconstructing the optimal prediction results of each sub-series. The monthly streamflow of the upper Fenhe River Basin, which is in the semi-humid and semi-arid climate zone, was selected as a case study. The results indicate that in comparison to both undecomposed and singly decomposed scenarios, the input–output bi-decomposed scenario more accurately identifies the input factors and constructs high-precision prediction models. The Nash–Sutcliffe efficiency (NSE) of both the S3-KELM and S3-HKELM models exceeds 0.85. Specifically, the S3-HKELM model demonstrates superior performance, capable of handling more complex inputs, with its NSE reaching up to 0.93. Importantly, meteorological and climate factors contribute to the accuracy of streamflow predictions across different scenarios.

Gilgit-Baltistan, Pakistan, is particularly susceptible to landslides due to various geological, tectonics, meteorological, and anthropogenic factors consequently. However, the persisting conundrum of landslide database/data imbalance stands as a formidable challenge within this domain. To better stabilize the objective of landslide prediction, stacking ensemble Machine Learning and Generative Adversarial Network (GAN) were applied, because previous research in this area has mostly been limited by a lack of data. GAN is employed to synthesize training samples, ensuring the creation of a balanced dataset. Stacking ensemble architecture involves two stages of learning: the first class of learners incorporates diverse machine learning algorithms, while, the second level logistic regression model integrates prediction based on the strong learner, thereby enhancing overall prediction performance. To investigate landslide susceptibility in District Chilas, Northern Pakistan, we employed optical remote sensing and introduced a GAN with a Multi-Layers Hybrid Model (MLHM). This study involved the preparation of a spatial database with a total of 106 landslides and ten major landslide factors. We utilized a hybrid ensemble model and compared its performance with different algorithms like Conventional Neural Network, Artificial Neural network, Decision Tree, K-Nearest Neighbouring, and Hybrid Model, achieving accuracies of 0.91, 0.92, 0.90, 0.89, and 0.93, respectively. this approach has with Hybrid architecture learning accuracy of 0.98. The GAN with MLHM developed improved landslide susceptibility assessment with cross-comparison of Persistent Scattered Interferometric Synthetic Aperture Radar (PS-InSAR) investigation to ensure the safe functioning of KKH.

The inhomogeneous Log-Gaussian Cox Process (LGCP) defines a flexible point process model for the analysis of spatial point patterns featuring inhomogeneity/spatial trend and aggregation patterns. To fit an LGCP model to spatial point pattern data and study the spatial trend, one could link the intensity function with continuous spatial covariates. Although non-continuous covariates are becoming more common in practice, the existing estimation methods so far only cover covariates in continuous form. As a consequence, to implement such methods, the non-continuous covariates are replaced by the continuous ones by applying some transformation techniques, which are many times problematic. In this paper, we develop a technique for inhomogeneous LGCP involving non-continuous covariates, termed piecewise constant covariates. The method does not require covariates transformation and likelihood approximation, resulting in an estimation technique equivalent to the one for generalized linear models. We apply our method for modeling COVID-19 transmission risk in East Java, Indonesia, which involves five piecewise constant covariates representing population density and sources of crowd. We outline that population density and industry density are significant covariates affecting the COVID-19 transmission risk in East Java.