Comparing state-of-the-art evolutionary multi-objective algorithms for long-term groundwater monitoring design

This study compares the performances of four state-of-the-art evolutionary multi-objective optimization (EMO) algorithms: the Non-Dominated Sorted Genetic Algorithm II (NSGAII), the Epsilon-Dominance Non-Dominated Sorted Genetic Algorithm II (ε-NSGAII), the Epsilon-Dominance Multi-Objective Evolutionary Algorithm (εMOEA), and the Strength Pareto Evolutionary Algorithm 2 (SPEA2), on a four-objective long-term groundwater monitoring (LTM) design test case. The LTM test case objectives include: (i) minimize sampling cost, (ii) minimize contaminant concentration estimation error, (iii) minimize contaminant concentration estimation uncertainty, and (iv) minimize contaminant mass estimation error. The 25-well LTM design problem was enumerated to provide the true Pareto-optimal solution set to facilitate rigorous testing of the EMO algorithms. The performances of the four algorithms are assessed and compared using three runtime performance metrics (convergence, diversity, and ε-performance), two unary metrics (the hypervolume indicator and unary ε-indicator) and the first-order empirical attainment function. Results of the analyses indicate that the ε-NSGAII greatly exceeds the performance of the NSGAII and the εMOEA. The ε-NSGAII also achieves superior performance relative to the SPEA2 in terms of search effectiveness and efficiency. In addition, the ε-NSGAII’s simplified parameterization and its ability to adaptively size its population and automatically terminate results in an algorithm which is efficient, reliable, and easy-to-use for water resources applications.     

A new epsilon-dominance hierarchical Bayesian optimization algorithm for large multiobjective monitoring network design problems

This study focuses on the development of a next generation multiobjective evolutionary algorithm (MOEA) that can learn and exploit complex interdependencies and/or correlations between decision variables in monitoring design applications to provide more robust performance for large problems (defined in terms of both the number of objectives and decision variables). The proposed MOEA is termed the epsilon-dominance hierarchical Bayesian optimization algorithm (ε$\epsilon$-hBOA), which is representative of a new class of probabilistic model building evolutionary algorithms. The ε$\epsilon$-hBOA has been tested relative to a top-performing traditional MOEA, the epsilon-dominance nondominated sorted genetic algorithm II (ε$\epsilon$-NSGAII) for solving a four-objective LTM design problem. A comprehensive performance assessment of the ε$\epsilon$-NSGAII and various configurations of the ε$\epsilon$-hBOA have been performed for both a 25 well LTM design test case (representing a relatively small problem with over 33 million possible designs), and a 58 point LTM design test case (with over 2.88×10¹⁷$2.88\times {10}^{17}$ possible designs). The results from this comparison indicate that the model building capability of the ε$\epsilon$-hBOA greatly enhances its performance relative to the ε$\epsilon$-NSGAII, especially for large monitoring design problems. This work also indicates that decision variable interdependencies appear to have a significant impact on the overall mathematical difficulty of the monitoring network design problem.     

Parallelization strategies for rapid and robust evolutionary multiobjective optimization in water resources applications

This study uses a formal metrics-based framework to demonstrate the Master–Slave (MS) and the Multiple-Population (MP) parallelization schemes for the Epsilon-Nondominated Sorted Genetic Algorithm-II (ε-NSGAII). The MS and MP versions of the ε-NSGAII generalize the algorithm’s auto-adaptive population sizing, ε-dominance archiving, and time continuation to a distributed processor environment using the Message Passing Interface. This study uses three test cases to compare the MS and MP versions of the ε-NSGAII: (1) an extremely difficult benchmark test function termed DTLZ6 drawn from the computer science literature, (2) an unconstrained, continuous hydrologic model calibration test case for the Leaf River near Collins, Mississippi, and (3) a discrete, constrained four-objective long-term groundwater monitoring (LTM) application. The MP version of the ε-NSGAII is more effective than the MS scheme when solving DTLZ6. Both the Leaf River and the LTM test cases proved to be more appropriately suited to the MS version of the ε-NSGAII. Overall, the MS version of the ε-NSGAII exhibits superior performance on both of the water resources applications, especially when considering its simplicity and ease-of-implementation relative to the MP scheme. A key conclusion of this study is that a simple MS parallelization strategy can exploit time-continuation and parallel speedups to dramatically improve the efficiency and reliability of evolutionary multiobjective algorithms in water resources applications.     

水库诱发地震地下水监测网建设的若干问题探讨

根据国内外已发生的水库诱发地震的基本特点,结合三峡工程诱发地震地下水监测网与金沙江下游水电工程地下水监测网建设的经验及我国地震地下水动态观测网建设与运行中得到的科学认识,探讨了水库诱发地震的地下水监测井网建设中的布网区及其范围、观测井间距确定、观测井位置选择、观测含水层选择与观测井深度和结构设计等若干技术问题.     

Impacts of riparian zone plant water use on temporal scaling of groundwater systems

In this work, we study groundwater system temporal scaling in relation to plant water use and near‐river‐stage fluctuations in riparian zones where phreatophytes exist. Using detrended fluctuation analysis (DFA), we investigate the influence of regular diurnal fluctuations due to phreatophyte water use on temporal scaling properties of groundwater level variations. We found that groundwater use by phreatophytes, at the field site on the Colorado River, USA, results in distinctive crossovers (slope changes when the plots are fitted with straight lines) in the logarithm plots of root‐mean‐square fluctuations of the detrended water level time series versus time scales of groundwater level dynamics. For groundwater levels monitored at wells close to the river, we identified one crossover at ～1 day in the scaling characteristics of groundwater level variations. When time scale exceeds 1 day, the scaling properties decrease from persistent to close to 1/f noise, where f is the frequency. For groundwater levels recorded at wells further away from the river, the slope of the straight line fit (i.e. scaling exponent) is smallest when the time scale is between 1 and 3 days. When the time scale is < 1 day, groundwater variations become persistent. When the time scale is between 1 and 3 days, the variations are close to white noise, but return to persistent when the time scale is > 3 days. Copyright © 2011 John Wiley & Sons, Ltd.     

Pumping optimization of coastal aquifers based on evolutionary algorithms and surrogate modular neural network models

Pumping optimization of coastal aquifers involves complex numerical models. In problems with many decision variables, the computational burden for reaching the optimal solution can be excessive. Artificial Neural Networks (ANN) are flexible function approximators and have been used as surrogate models of complex numerical models in groundwater optimization. However, this approach is not practical in cases where the number of decision variables is large, because the required neural network structure can be very complex and difficult to train. The present study develops an optimization method based on modular neural networks, in which several small subnetwork modules, trained using a fast adaptive procedure, cooperate to solve a complex pumping optimization problem with many decision variables. The method utilizes the fact that salinity distribution in the aquifer, depends more on pumping from nearby wells rather than from distant ones. Each subnetwork predicts salinity in only one monitoring well, and is controlled by relatively few pumping wells falling within certain control distance from the monitoring well. While the initial control area is radial, its shape is adaptively improved using a Hermite interpolation procedure. The modular neural subnetworks are trained adaptively during optimization, and it is possible to retrain only the ones not performing well. As optimization progresses, the subnetworks are adapted to maximize performance near the current search space of the optimization algorithm. The modular neural subnetwork models are combined with an efficient optimization algorithm and are applied to a real coastal aquifer in the Greek island of Santorini. The numerical code SEAWAT was selected for solving the partial differential equations of flow and density dependent transport. The decision variables correspond to pumping rates from 34 wells. The modular subnetwork implementation resulted in significant reduction in CPU time and identified an even better solution than the original numerical model.     

Evaluation of hydrologic data obtained from a local groundwater monitoring network in a metropolitan city,Korea

In the late 1980s, dramatic increases in water use caused over‐exploitation of groundwater resources and deterioration of water quality in Seoul metropolitan city. To monitor changes in quantity of groundwater resources and their quality, the metropolitan government established a local groundwater monitoring network in 1997 consisting of 119 monitoring wells. Groundwater resources in the urban area were affected by various human activities, including underground construction such as subways, pumping for public or private water use, leaky sewer systems and pavements. The variation patterns of the groundwater levels were mainly classified into four types, reflecting natural recharge due to rainfall events during the wet season, artificial recharge from leaky sewer or water supply systems, and heavy groundwater pumping for drainage or flood control purposes at underground construction sites. Significantly decreasing trends of groundwater levels in the suburbs of Seoul indicate groundwater use for various agricultural activities. Subway construction lowered the water level by an average of 25 m. Electrical conductivity values showed a wide range, from 100 to 1800 µS/cm (mean 470 µS/cm). Groundwater temperature generally showed a stable pattern, except for some sensitive increases at relatively shallow monitoring wells. Detailed analysis of the monitored groundwater data would provide some helpful implications for optimal and efficient management of groundwater resources in this metropolitan city. Copyright © 2005 John Wiley & Sons, Ltd.     

Spatial and temporal analysis of groundwater recharge with application to sampling design

A numerical experiment of flow in variably saturated porous media was performed in order to evaluate the spatial and temporal distribution of the groundwater recharge at the phreatic surface for a shallow aquifer as a function of the input rainfall process and soil heterogeneity. The study focused on the groundwater recharge which resulted from the percolation of the excess rainfall for a 90-days period of an actual precipitation record. Groundwater recharge was defined as the water flux across the moving phreatic surface. The observed spatial non-uniformity of the groundwater recharge was caused by soil heterogeneity and is particularly pronounced during the stage of recharge peak (substantial percolation stage). During that stage the recharge is associated with preferential flow paths defined as soil zones of locally higher hydraulic conductivity. For the periods of low percolation intensity the groundwater recharge was exhibiting more uniform spatial characteristics. The temporal distribution of the recharge was found to be a function of the frequency and intensity of the rainfall events. Application of sampling design demonstrates the joint influence of the spatial and temporal recharge variability on the cost-effective monitoring of groundwater potentiometric surfaces.     

Evaluation of long-term groundwater level data in regular monitoring wells,Barka, Sultanate of Oman

A detailed investigation was carried out to evaluate long-term groundwater level fluctuation in regular monitoring wells constructed by the Ministry of Water Resources in Barka, Sultanate of Oman. For this study, groundwater level data for 71 wells and rainfall data from six stations were collected from 1984 to 2003 and analysed. Based on long-term water level fluctuation, groundwater wells are classified into three groups. In group 1, water level shows a long-term cyclic trend without yearly fluctuation whereas in group 2 the water level declined continuously until 1995 followed by a constant water level. In group 3, water level decreases continuously throughout the study periods with rapid annual cyclic variation. Group 1 wells show high water-level fluctuations (5 to 10 m) and seem to be regulated by discharge (lateral flow) from this aquifer and recharge from the adjacent Jabal Akhdar mountainous region. Constant trend in water level after 1995 in group 2 wells illustrates the advancement of saline–fresh water interface to the inland due to heavy pumping which is justified by higher electrical conductivity and Cl/HCO₃ ratio. In group 3 wells, the water level dropped continuously due to overabstraction by agricultural farms and human settlements. In addition, wells existing near the recharge dams express the influences of recharge dams and rainfall, and exhibit high water-level fluctuations during heavy rainfall periods. The long-term regional variation indicates that water level drops continuously in the coastal and central parts of the study region. Linear regression analysis revealed that the decline in water level is 0·3–0·4 m year⁻¹ near the coastal and central parts of the study area and is almost constant in the remaining area. We conclude that the contribution of man-made activities on groundwater level is well compared with natural factors. Copyright © 2007 John Wiley & Sons, Ltd.     

Quantifying uncertainty of the semivariogram of transmissivity of an existing groundwater monitoring network

A method is presented for quantifying the uncertainty of the semivariogram of transmissivity and determining the required number of measurements. In this method, the estimated semivariogram and its 95% confidence limits are first determined from a finite number of measurements. The uncertainty of the estimated semivariogram is then quantified using the random field simulation technique. For a given value of the quantitative index of uncertainty, the required number of measured data can finally be obtained. Actual transmissivity data of an existing groundwater monitoring network are used in the application of the proposed method. The required numbers of measurements of transmissivity for four different values of the quantitative index of uncertainty are provided, from which reliable semivariograms of the transmissivity can be obtained. Copyright © 2004 John Wiley & Sons, Ltd.     

Role of reservoir regulation and groundwater feedback in a simulated ground-soil-vegetation continuum: A long-term regional scale analysis

The regional terrestrial water cycle is strongly altered by human activities. Among them, reservoir regulation is a way to spatially and temporally allocate water resources in a basin for multi-purposes. However, it is still not sufficiently understood how reservoir regulation modifies the regional terrestrial- and subsequently, the atmospheric water cycle. To address this question, the representation of reservoir regulation into the terrestrial component of fully coupled regional Earth system models is required. In this study, an existing process-based reservoir network module is implemented into NOAH-HMS, that is, the terrestrial component of an atmospheric–hydrologic modelling system, namely, the WRF-HMS. It allows to quantitatively differentiate role of reservoir regulation and of groundwater feedback in a simulated ground-soil-vegetation continuum. Our study focuses on the Poyang Lake basin, where the largest freshwater lake of China and reservoirs of different sizes are located. As compared to streamflow observations, the newly extended NOAH-HMS slightly improves the streamflow and streamflow duration curves simulation for the Poyang Lake basin for the period 1979–1986. The inclusion of reservoir regulation leads to major changes in the simulated groundwater recharges and evaporation from reservoirs at local scale, but has minor effects on the simulated soil moisture and surface runoff at basin scale. The performed groundwater feedback sensitivity analysis shows that the strength of the groundwater feedback is not altered by the consideration of reservoir regulation. Furthermore, both reservoir regulation and groundwater feedback modify the partitioning of the simulated evapotranspiration, thus affecting the atmospheric water cycle in the Poyang Lake region. This finding motivates future research with our extended fully coupled atmospheric–hydrologic modelling system by the community.     

Assessment of groundwater recharge processes through karst vadose zone by cave percolation monitoring

Recharge processes of karst aquifers are difficult to assess given their strong heterogeneity and the poorly known effect of vadose zone on infiltration. However, recharge assessment is crucial for the evaluation of groundwater resources. Moreover, the vulnerability of karst aquifers depends on vadose zone behaviour because it is the place where most contamination takes place. In this work, an in situ experimental approach was performed to identify and quantify flow and storage processes occurring in karst vadose zone. Cave percolation monitoring and dye tracing were used to investigate unsaturated zone hydrological processes. Two flow components (diffuse and quick) were identified and, respectively, account for 66% and 34% of the recharge. Quickflow was found to be the result of bypass phenomenon in vadose zone related to water saturation. We identify the role of epikarst as a shunting area, most of the storage in the vadose zone occurring via the diffuse flow component in low permeability zones. Relationship between rainfall intensity and transit velocity was demonstrated, with 5 times higher velocities for the quick recharge mode than the diffuse mode. Modelling approach with KarstMod software allowed to simulate the hybrid recharge through vadose zone and shows promising chances to properly assess the recharge processes in karst aquifer based on simple physical models.     

Visual analytics clarify the scalability and effectiveness of massively parallel many-objective optimization: A groundwater monitoring design example

In this study, we contribute a comprehensive framework for simultaneously assessing solution quality and scalability for massively parallel multiobjective evolutionary algorithm (MOEA)-based search using a highly challenging optimization—assimilation application. Visual analytics are used to evaluate how changes in search metric performance relate to actual decision relevant changes in the Pareto approximate set. The application focuses on a four objective groundwater monitoring application in which parallel scalability is tested across compute core counts ranging from 64 to a maximum of 8192. This study demonstrates that parallel search performance must be assessed in terms of how well speedup is exploited to improve the quality of search results and that solely focusing on differences in computational time can be deceptive. Our results demonstrate how visualization can clarify when an MOEA’s search shifts from “translating” the approximation set to “diversifying” its coverage over the extent of the objectives. This is an important observation. If shorter parallel run durations are required, the rapid early translation of the set may yield a reasonable approximation of the Pareto approximate set where further search is unnecessary. Although a groundwater application is used to demonstrate our parallelization, the visual analytics and metrics utilized to characterize the parallel scalability of MOEA-based search are broadly applicable in water resources and beyond.     

Ecosystem response to regulatory and management actions: The southern California experience in long-term monitoring

Billions of dollars have been invested over the past 35 years in reducing pollutant emissions to coastal environments. Evaluation of the effectiveness of this investment is hampered by the lack of long-term consistent data. A rare opportunity exists in southern California to evaluate the effectiveness of management actions by analyzing long-term monitoring of effluent, sediment, benthos, and fish and comparing this trend data to periodic regional surveys of environmental condition. In this paper, we ask the question “have improvements in effluent quality in response to environmental regulation translated into improvements in the receiving environment?” Results indicate that management actions directed at reducing mass emissions from wastewater treatment plants (POTWs) have resulted in substantial improvement in aquatic communities. However, the magnitude and timing of response varies by indicator suggesting that use of multiple assessment endpoints is necessary to adequately interpret trends. Reductions in the effect of POTW effluent have allowed managers to shift resources to address other contaminant sources such as stormwater and resuspension of legacy pollutants.     

结合长时序遥感监测和机器学习算法预测藻类增殖风险时空变化

饮用水源地藻类增殖监测和预测对于改善水生态系统环境和保护人类健康具有重要意义。利用多源遥感数据能够获取高时空分辨率的藻类动态信息,结合长时序遥感监测与机器学习算法能够适应藻类生长复杂的影响机制和非线性特征,实现藻类增殖风险时空变化的预测。本文利用Landsat与MODIS长时间序列卫星遥感数据,采用FAI与NDVI两种方法提取2000—2020年丹江口水库藻类浓度的时空变化信息,在此基础上分析藻类增殖对气象因子(气温、气压、相对湿度、风速和累计日照时间)的时间滞后效应。利用支持向量机、朴素贝叶斯与随机森林3种机器学习算法预测藻类增殖风险,并对3种算法的预测性能进行评价和对比。结果表明:丹江口水库藻类浓度年际变化呈现出先增后降的趋势,呈现出明显的季节性周期变化,春末夏初是藻类快速增殖时期。空间上入库支流和库湾等局部地区藻类浓度相对较高,为藻类增殖高风险区,丹江口水库藻类增殖风险预测模型能够较为准确地确定藻类增殖高风险区位置并反映短期内的空间变化情况,3种算法的预测结果呈现出整体上的一致性,其中支持向量机与朴素贝叶斯算法表现出更高的精度,提前4～5天是最佳预测时间。     

Improving the reliability of groundwater monitoring networks using combined numerical,geostatistical and neural network-based simulation models

ABSTRACT

A new methodology is proposed for improving the accuracy of groundwater-level estimations and increasing the efficiency of groundwater-level monitoring networks. Three spatio-temporal (S-T) simulation models, numerical groundwater flow, artificial neural network and S-T kriging, are implemented to simulate water-table level variations. Individual models are combined using model fusion techniques and the more accurate of the individual and combined simulation models is selected for the estimation. Leave-one-out cross-validation shows that the estimation error of the best fusion model is significantly less than that of the three individual models. The selected fusion model is then considered for optimal S-T redesign of the groundwater monitoring network of the Dehgolan Plain (Iran). Using a Bayesian maximum entropy interpolation technique, soft data are included in the geostatistical analyses. Different scenarios are defined to incorporate economic considerations and different levels of precision in selecting the best monitoring network; a network of 37 wells is proposed as the best configuration. The mean variance estimation errors of all scenarios decrease significantly compared to that of the existing monitoring network. A reduction in equivalent uniform annual costs of different scenarios is achieved.     

A field demonstration of the entropy-weighted fuzzy DRASTIC method for groundwater vulnerability assessment

Abstract

Groundwater vulnerability assessment based on the DRASTIC index has been widely used since the 1980s to map potential risks of groundwater contamination. However, its applicability and usefulness are affected by two uncertain and subjective factors. One is the discretization of continuous input variables and the other is the assignment of different weights to the index variables. In this study, an entropy-weighted fuzzy-optimization approach was developed to augment and improve the classic DRASTIC method by reducing the uncertainties associated with variable discretization and weight assignment. The modified DRASTIC method was applied to a study site in Shandong, north China. The entropy-weighted fuzzy-optimization approach is shown to provide a more rigorous delineation of the relative vulnerability distribution. Meanwhile, the new approach does not require the use of more parameters. The results suggest that this approach significantly improves and enhances the ability of the classic DRASTIC method in a more systematic and rigorous way.

Editor D. Koutsoyiannis

Citation Yu, C., Zhang, B.X., Yao, Y.Y., Meng, F.H., and Zheng, C.M., 2012. A field demonstration of the entropy-weighted fuzzy DRASTIC method for groundwater vulnerability assessment. Hydrological Sciences Journal, 57 (7), 1420–1432.     

宁波市地下水位动态与地面沉降预测分析

为了准确预测分析宁波市地下水位动态与地面沉降的发展趋势,建立了宁波市第四纪松散沉积层孔隙地下水流三维数值模拟模型和地面沉降与地下水位多元线性回归模型,预测了2009年底到2020年底的逐月地下水位动态和逐年地面沉降量的变化特征.结果表明,从2013年起,除山区沟谷孔隙潜水地下水位降落漏斗逐渐扩大外,其余孔隙水的地下水流场基本趋于稳定,地下水位年际变化很小,年地面沉降量也逐渐变小,由2012年的5.62 mm/a逐渐下降到2020年的5.54 mm/a,由地下水位下降引起的地面沉降基本得到控制.     

云南地区地震地下水宏观异常统计分析

对1976—2001年云南38组M≥5地震震例中地下水宏观异常进行统计分析,得到这些震例地下水宏观异常的共同特征;再对2011—2021年云南各地州、市通过地震行业内网上报的755项宏观异常中705项地下水宏观异常进行梳理;参照震例中地下水宏观异常的共同特征对云南M≥5地震与地下水宏观异常间的关系进行分析。结果表明,云南地区地下水宏观异常出现在震前3个月内的、出现在极震区内的比例均为100%。     

Quantitative analysis of the driving factors for groundwater resource changes in arid irrigated areas

How to quantify the impact of climate change and human activities on groundwater is not only a hot topic of current research but also a key point of water resource management in arid irrigated areas. Therefore, this paper analyzes the changes in the trends of land use, climate, and groundwater extraction in the Yanqi Basin in recent years and uses the distributed hydrological model MIKE-SHE to quantitatively analyze the impacts of these three factors on groundwater resources. The results show that: 1. The Nash coefficients of the simulated and observed groundwater levels during the verification period are 0.84, 0.79 and 0.76; the correlation coefficient between the simulated and observed soil moisture is 0.86. Although there are some uncertainties in the simulation, the results prove that the model can be used to simulate arid irrigated areas. 2. The effects of these three factors on groundwater levels are 5, 12.5 and 82.5%, respectively, and have caused the regional average groundwater level to decrease by a maximum of 0.07, 0.23 and 1.79 m, respectively. The effects of these three factors on the interactions between surface water and groundwater were 7.04, 3.63 and 89.33%. Groundwater extraction has become the main influencing factor of regional groundwater resources changes due to its more direct influence. 3. The influence of groundwater extraction has a strong spatial distribution characteristic and 10% of the study area has been greatly impacted by the groundwater extraction. Base on the above results, integrating multidisciplinary knowledge to establish the relationship between ecological environment and groundwater changes can provide strategies for the sustainable development of groundwater.