人工神经网络技术在水文地质学中的应用潜力初探

在许多水文地质问题中,多因素且非线性的影响常使传统的集中参数随机模型或分布参数确定性数值模型的方法难以对其作出符合实际的评价与预测。本文从几个典型的水文地质问题入手,利用人工神经网络技术的高度自组织、自适应与自学习能力和分类计算能力,对这些问题的解决进行了系统的ＢＰ网分析。结果表明,人工神经网络的应用可有效减少人为的主观臆断性,其训练识别的结果更符合实际,效果令人满意,因此具有十分广阔的应用前景。     

Spatial and temporal monthly precipitation forecasting using wavelet transform and neural networks,Qara-Qum catchment,Iran

This paper aims to provide a spatial and temporal analysis to prediction of monthly precipitation data which are measured at irregularly spaced synoptic stations at discrete time points. In the present study, the rainfall data were used which were observed at four stations over the Qara-Qum catchment, located in the northeast of Iran. Several models can be used to spatially and temporally predict the precipitation data. For temporal analysis, the wavelet transform with artificial neural network (WTANN) framework combines with the wavelet transform, and an artificial neural network (ANN) is used to analyze the nonstationary precipitation time-series. The time series of dew point, temperature, and wind speed are also considered as ancillary variables in temporal prediction. Furthermore, an artificial neural network model was used for comparing the results of the WTANN model. Therefore, four models were developed, including WTANN and ANN with and without ancillary data. Several statistical methods were used for comparing the results of the temporal analysis. It was evident that at three of the four stations, the WTANN models were more effective than the ANN models, and only at one station, the ANN model with ancillary data had better performance than the WTANN model without ancillary data. The values of correlation coefficient and RMSE for WTANN model with ancillary data for the validation period at Mashhad station which showed the best results were equal to 0.787 and 13.525 mm, respectively. Finally, an artificial neural network model was used as an alternative interpolating technique for spatial analysis.     

Determination of the rate and volume of leakage using the slopes of time-drawdown data

Quantification of leakage is very important in the selection and design of the remediation systems of leaky aquifers that receive contaminated leakage. This is an approach for the calculation of leakage using only two slopes of time-drawdown data. These slopes represent before and after the start of leakage, and are applied to four examples. Results generally agree with those determined by the Hantush approach. Comparison of the two approaches, however, shows that the Hantush approach quantifies leakage using three aquifer parameters (transmissivity, storativity, and leakage factor), the value of which depend on the pumping test method used; it assumes constant hydraulic head in the aquifer supplying leakage, which may not be valid under field conditions; and it ignores differences between the viscosities of the leakage water and the aquifer water, which influence the leakage rate. The proposed approach is free from all three limitations.     

Water sustainability model for estimation of groundwater availability in Kemuning district,Riau-Indonesia

There are rising interests in the utility of groundwater in various aspects,which is capable of triggering problematic issues.The excessive exploitation for anthropologic uses,without regards to aquifer capacity,will decreases the water table as well as capacity of groundwater in the aquifer.This research was aimed to provide aquifer model of underground water by consideration of various environmental factors,with the propensity of being modeled,in an attempt to predict groundwater conditions in subsequent years.The purpose of this research was to forecast water requirements,availability,as well as three-dimensional model of groundwater depth in Kemuning,Indragiri Hilir Regency-Indonesia between 2015 and 2022.Furthermore,various environmental factors,from aquifer profiles to anthropologic demand,are taken into account in the evaluated model,including water requirements,encompassing recharge and aquifer parameters,which consists of storativity and transmissivity.From anthropologic side are domestic requirements,trade,public facilities,agriculture,and livestock.The results show that groundwater availability in Kemuning is to be safe condition,and average difference is 1.06×108 m3/yr.The coefficient of storativity and transmissivity are 16.514 m2/day and 9897.26 m2/day,respectively,while the average depth was recorded as 2.8965 m to 10.4927 m.     

Effects of parameter error on Cooper-Jacob drawdown estimates

Parameters employed in the Cooper-Jacob equation to describe drawdown are transmissivity, storativity, radial distance, time and pumping rate. An approach is described for quantifying how error or uncertainty in any one of the parameters used causes error in estimated drawdown. Dimensionless fractional error in estimated drawdown is expressed quantitatively as a function of (1) dimensionless fractional error of a given parameter, and (2) dimensionless argument of the well function, u. Fractional error in estimated drawdown is a linear function of fractional error in pumping rate and, for any given value of u, a nonlinear function of fractional error in transmissivity, storativity, radial distance or time. Fractional error in estimated drawdown for a given fractional parameter error varies considerably between parameters. The greatest sensitivity is for transmissivity and flow rate. Sensitivity is less for radial distance and time, and even less for storativity. The magnitude of the fractional error in drawdown may be affected by the sign of the fractional parameter error.     

A comparative study of shallow groundwater level simulation with WA–ANN and ITS model in a coastal island of south China

Accurate and reliable prediction of shallow groundwater level is a critical component in water resources management. Two nonlinear models, WA–ANN method based on discrete wavelet transform (WA) and artificial neural network (ANN) and integrated time series (ITS) model, were developed to predict groundwater level fluctuations of a shallow coastal aquifer (Fujian Province, China). The two models were testified with the monitored groundwater level from 2000 to 2011. Two representative wells are selected with different locations within the study area. The error criteria were estimated using the coefficient of determination (R ²), Nash–Sutcliffe model efficiency coefficient (E), and root-mean-square error (RMSE). The best model was determined based on the RMSE of prediction using independent test data set. The WA–ANN models were found to provide more accurate monthly average groundwater level forecasts compared to the ITS models. The results of the study indicate the potential of WA–ANN models in forecasting groundwater levels. It is recommended that additional studies explore this proposed method, which can be used in turn to facilitate the development and implementation of more effective and sustainable groundwater management strategies.     

Prediction of relative crest settlement of concrete-faced rockfill dams analyzed using an artificial neural network model

A neural network model has been developed for the prediction of relative crest settlement (RCS) of concrete-faced rockfill dams (CFRDs) using 30 databases of field data from seven countries (of which 21 were used for training and 9 for testing). The settlement values predicted using the optimum artificial neural network (ANN) model are in good agreement with these field data. A database prepared from reported crest settlement values of CFRDs after construction was used to train the ANN model to predict the RCS. It is demonstrated here that the model is capable of predicting accurately the relative crest settlement of CFRDs and is potentially applicable for general usage with knowledge of the three basic properties of a dam (void ratio, e; height, H; and vertical deformation modulus, E_V).

The performance of the new ANN model is compared with that of conventional methods based on the Clements theory and also with that of a proposed equation derived from the field data. The comparison indicates that the ANN model has strong potential and offers better performance than conventional methods when used as a quick interpolation and extrapolation tool. The conventional calculation model was proposed based on the fixed connection weights and bias factors of the optimum ANN structure. This method can support the dam engineer in predicting the relative crest settlement of a CFRD after impounding.     

Forecasting of groundwater level in hard rock region using artificial neural network

In hardrock terrain where seasonal streams are not perennial source of freshwater, increase in ground water exploitation has already resulted here in declining ground water levels and deteriorating its’ quality. The aquifer system has shown signs of depletion and quality contamination. Thus, to secure water for the future, water resource estimation and management has urgently become the need of the hour. In order to manage groundwater resources, it is vital to have a tool to predict the aquifer response for a given stress (abstraction and recharge). Artificial neural network (ANN) has surfaced as a proven and potential methodology to forecast the groundwater levels. In this paper, Feed-Forward Network based ANN model is used as a method to predict the groundwater levels. The models are trained with the inputs collected from field and then used as prediction tool for various scenarios of stress on aquifer. Such predictions help in developing better strategies for sustainable development of groundwater resources.     

利用不对称的地下水位潮汐波动确定滨海含水层参数

在滨海地区,地下水水位受潮汐波动影响较大,使得传统的抽水试验、水位回复试验等方法确定含水层参数存在困难且花费较大。通过对广西北海市滨海含水层地下水位动态资料进行分析,发现其上升段和下降段是不对称的。基于海岸带承压含水层正弦潮汐波的传播理论,提出了确定含水层参数的分段法,并与振幅衰减法和滞后时间法进行对比,各种方法求出的储水系数与导水系数之比(S/T)很接近,说明分段法是有效的。对于北海市滨海含水层,上升段求出的S/T值比下降段要大,其成因机理还有待进一步分析。     

Inversion of quasi-3D DC resistivity imaging data using artificial neural networks

The objective of this paper is to investigate the applicability of artificial neural networks in inverting quasi-3D DC resistivity imaging data. An electrical resistivity imaging survey was carried out along seven parallel lines using a dipole-dipole array to confirm the validation of the results of an inversion using an artificial neural network technique. The model used to produce synthetic data to train the artificial neural network was a homogeneous medium of 100Ωm resistivity with an embedded anomalous body of 1000Ωm resistivity. The network was trained using 21 datasets (comprising 12159 data points) and tested on another 11 synthetic datasets (comprising 6369 data points) and on real field data. Another 24 test datasets (comprising 13896 data points) consisting of different resistivities for the background and the anomalous bodies were used in order to test the interpolation and extrapolation of network properties. Different learning paradigms were tried in the training process of the neural network, with the resilient propagation paradigm being the most efficient. The number of nodes, hidden layers, and efficient values for learning rate and momentum coefficient have been studied. Although a significant correlation between results of the neural network and the conventional robust inversion technique was found, the ANN results show more details of the subsurface structure, and the RMS misfits for the results of the neural network are less than seen with conventional methods. The interpreted results show that the trained network was able to invert quasi-3D electrical resistivity imaging data obtained by dipole-dipole configuration both rapidly and accurately.     

基于模糊人工神经网络识别的水质评价模型

人工神经网络和模糊识别理论作为模拟生物体的信息处理系统,在实践中得到广泛的应用,且各有所长,将二者相结合,构造出模糊人工神经识别网络,从而使识别系统的柔性处理能力得到很大提高.阐述了模型和方法,将其用于长江支流沱江枯水期的水质综合评价,结果表明,模糊人工神经网络综合评价具有客观性和实用性.     

Multi-observation well aquifer test case study: is recovery coincident with the cessation of pumping?

Due to increases in water demand, the City of Kenedy, TX, USA must expand their small drinking water supply in the Gulf Coast aquifer system. Groundwater wells owned by the City of Kenedy, Karnes County, TX were examined to estimate properties for the Jasper aquifer. Conditions of four wells were assessed, after which two wells were rehabilitated and used as pumping wells in aquifer tests. Aquifer tests show that recovery in observation wells was not coincident with the cessation of pumping. Post-pumping data were selectively excluded so that only recovery data were used for analyses. Transmissivity for the Jasper aquifer ranges from 102 to 242 m² d^?1, and storativity ranges from 6.9E?05 to 3.3E?04. Transmissivity computed from recovery data was approximately 25 % higher than transmissivity computed from time-drawdown data. Field measured specific capacities and drawdowns were compared to theoretical specific capacities and drawdowns to calculate pumping well efficiencies in the range of 52.2–99.4 %. This study indicates that water demand for the City of Kenedy could be met by incorporating the tested wells into the water supply system. Future studies should be designed to estimate groundwater recharge rates and a complete water balance for computing a sustainable maximum annual yield.     

Dimensionless straight line fitting method for hydrogeological parameter determination

Aquifers may have alluvium deposits, weathered layers, fractured zones, and karstic formations separately or in mixture forms. Such geological configurations do not allow classical aquifer test applicability, due to a set of underlying assumptions that are not usually valid in nature. In practice, the Jacob straight line method is the most commonly used approach for aquifer parameter determinations. Constant transmissivity and storativity estimations depend on large time-drawdown plots on semilogarithmic paper as a straight line. A common mistake is that the appearance of a general trend as a straight line on semilogarithmic paper is taken as guaranteed for the application of Jacob method. Since Jacob straight line is the large time extension of Theis type curve, there is only one straight line on the semilogarithmic paper that can represent Jacob method, which is based on the assumption that the aquifer is porous and homogeneous. In such a case, the Jacob method slope should equal to 2.3, which shows its validity. Otherwise, a modification of Jacob method is suggested in this paper. The basis of the methodology is a dimensionless type straight line approach for the aquifer parameter assessment. Its application is presented for aquifer test data from Oude Korendjik porous medium aquifer data. The application results indicate that the classical Jacob straight line method might not be valid without a preliminary check. The dimensionless reevaluation of existing data helps to check the validity. The necessary formulations for the modification of the classical straight line method are derived, which reduce to classical Jacob method for a specific set of parameters.     

Sensitivity analysis of groundwater level in Jinci Spring Basin (China) based on artificial neural network modeling

Jinci Spring in Shanxi, north China, is a major local water source. It dried up in April 1994 due to groundwater overexploitation. The groundwater system is complex, involving many nonlinear and uncertain factors. Artificial neural network (ANN) models are statistical techniques to study parameter nonlinear relationships of groundwater systems. However, ANN models offer little explanatory insight into the mechanisms of prediction models. Sensitivity analysis can overcome this shortcoming. In this study, a back-propagation neural network model was built based on the relationship between groundwater level and its sensitivity factors in Jinci Spring Basin; these sensitivity factors included precipitation, river seepage, mining drainage, groundwater withdrawals and lateral discharge to the associated Quaternary aquifer. All the sensitivity factors were analyzed with Garson’s algorithm based on the connection weights of the neural network model. The concept of “sensitivity range” was proposed to describe the value range of the input variables to which the output variables are most sensitive. The sensitivity ranges were analyzed by a local sensitivity approach. The results showed that coal mining drainage is the most sensitive anthropogenic factor, having a large effect on groundwater level of the Jinci Spring Basin.     

Prediction of pile settlement using artificial neural networks based on standard penetration test data

In recent years artificial neural networks (ANNs) have been applied to many geotechnical engineering problems with some degree of success. With respect to the design of pile foundations, accurate prediction of pile settlement is necessary to ensure appropriate structural and serviceability performance. In this paper, an ANN model is developed for predicting pile settlement based on standard penetration test (SPT) data. Approximately 1000 data sets, obtained from the published literature, are used to develop the ANN model. In addition, the paper discusses the choice of input and internal network parameters which were examined to obtain the optimum model. Finally, the paper compares the predictions obtained by the ANN with those given by a number of traditional methods. It is demonstrated that the ANN model outperforms the traditional methods and provides accurate pile settlement predictions.     

基于水力层析法的某煤矿承压含水层叠加放水试验分析

水文地质参数是衡量含水层富水性的重要指标,其空间分布特征的准确推算,对于矿井突水灾害预报与防治具有积极意义。传统放水试验分析方法得到的是等效水文地质参数,不能准确刻画含水层的空间异质性。基于随时间变化的水头数据与水文地质参数的互相关分析,将水力层析法应用在陕西榆林柠条塔煤矿井下承压含水层叠加放水试验数据分析中,获得了研究区水文地质参数的空间分布特征。结果表明,工作面南北富水性差异大,涌水区位于强导水带上;从涌水区往北,导水系数和储水系数整体上逐渐减小;南部导水系数和储水系数均较大,属于富水异常区。研究表明,水力层析法是一种有效的非均质含水层参数识别的新技术,将井下放水视为针对含水层的刺激源,结合水头与水文地质参数的互相关分析,联合多个观测孔的水头响应数据,反演刻画研究区域的主要地质结构特征。在矿井水文地质条件分析中,预先采用水力层析法识别富水异常区域,可以有效降低突水事故风险。     

Neural network prediction of nitrate in groundwater of Harran Plain, Turkey

Monitoring groundwater quality by cost-effective techniques is important as the aquifers are vulnerable to contamination from the uncontrolled discharge of sewage, agricultural and industrial activities. Faulty planning and mismanagement of irrigation schemes are the principle reasons of groundwater quality deterioration. This study presents an artificial neural network (ANN) model predicting concentration of nitrate, the most common pollutant in shallow aquifers, in groundwater of Harran Plain. The samples from 24 observation wells were monthly analysed for 1 year. Nitrate was found in almost all groundwater samples to be significantly above the maximum allowable concentration of 50 mg/L, probably due to the excessive use of artificial fertilizers in intensive agricultural activities. Easily measurable parameters such as temperature, electrical conductivity, groundwater level and pH were used as input parameters in the ANN-based nitrate prediction. The best back-propagation (BP) algorithm and neuron numbers were determined for optimization of the model architecture. The Levenberg–Marquardt algorithm was selected as the best of 12 BP algorithms and optimal neuron number was determined as 25. The model tracked the experimental data very closely (R = 0.93). Hence, it is possible to manage groundwater resources in a more cost-effective and easier way with the proposed model application.     

Positive and negative impacts of longwall mine subsidence on a sandstone aquifer

 Subsidence due to longwall underground coal mining changes the hydraulic properties, heads, yields, and in some cases the groundwater chemistry of overlying bedrock aquifers. A 7-year study of a sandstone aquifer overlying an active longwall mine in Illinois has supported a comprehensive model of these impacts. Subsidence caused increases in permeability and storativity over the longwall panel. These changes initially caused a major decline in water levels in the sandstone, but the aquifer recovered slightly within a few months and fully within several years after mining. The enhanced hydraulic properties combined with potentiometric recovery resulted in a zone of greater well yield. However, at sites with very poor transmissivity and inadequate recharge pathways, recovery may not occur. Also, at the study site, the physical enhancement was accompanied by a deterioration in groundwater quality from slightly brackish, sodium bicarbonate water to more brackish water with increased sulfate levels. Received: 17 March 1997 · Accepted: 9 September 1997     

Inverse groundwater modelling in the Willunga Basin,South Australia

A new inverse technique for modelling groundwater flow, based on a functional minimization technique, has been used to calibrate a groundwater flow model of a subregion of the Port Willunga aquifer within the Willunga Basin in South Australia. The Willunga Basin is the location of extensive viticulture, irrigated primarily by groundwater, the levels and quality of which have declined significantly over the last 40 years. The new method is able to generate estimates of transmissivity, storativity and groundwater recharge over the whole subregion as a time-varying continuous surface; previous methods estimate local discrete parameter values at specific times. The new method has also been shown to produce accurate head values for the subregion and very good estimates of groundwater recharge. Its ultimate goal will be to provide a new and invaluable tool for significantly improved groundwater resource management. Supported in part by US National Science Foundation grants, DMS-0107492 and DMS-0079478.     

Neural network based prediction of ground surface settlements due to tunnelling

Ground surface settlement due to tunnel excavation varies in magnitude and trend depending on several factors such as tunnel geometry, ground conditions, etc. Although there are several empirical and semi-empirical formulae available for predicting ground surface settlement, most of these do not simultaneously take into consideration all the relevant factors, resulting in inaccurate predictions. In this study, an artificial neural network (ANN) is incorporated with '113' of monitored field results to predict surface settlement for a tunnel site with prescribed conditions. To achieve this, a standard format (a protocol) for a database of monitored field data is first proposed and then used for sorting out a variety of monitored data sets available in KICT. Using the capabilities of pattern recognition and memorization of the ANN, an attempt is made to capture the rich physical characteristics smeared in the database and at the same time filter inherent noise in the monitored data. Here, an optimal neural network model is suggested through preliminary parametric studies. It is shown that preliminary studies for generating an optimal ANN under given training data sets are necessary because no analytical method for this purpose is available to date. In addition, this study introduces a concept of relative strength of effects (RSE) [Yang Y, Zhang Q. A heirarchical analysis for rock engineering using artificial neural networks. Rock Mechanics and Rock Engineering 1997; 30(4): 207–22] in sensitivity analysis for various major factors affecting the surface settlement in tunnelling. It is seen in some examples that the RSE rationally enables us to recognize the most significant factors of all the contributing factors. Two verification examples are undertaken with the trained ANN using the database created in this study. It is shown from the examples that the ANN has adequately recognized the characteristics of the monitored data sets retaining a generality for further prediction. It is believed that an ANN based hierarchical prediction procedure shown in this paper can be further employed in many kinds of geotechnical engineering problems with inherent uncertainties and imperfections.