多年平均气温空间化BP神经网络模型的模拟分析

气温数据空间化是插补无站地区温度、使气温数据便于综合分析的重要技术手段.理想情况下,气温的空间化分布受经度、纬度和海拔高度的影响,呈现规律性的空间分布态势.但是,各种微观因子如坡度、坡向、地形起伏、地表覆被等的存在,在一定程度上扰乱并弱化了这种规律性的分布态势.本文基于Matlab平台,利用BP神经网络研究了多年平均气...     

Predicting and forecasting flow discharge at sites receiving significant lateral inflow

Two models, one linear and one non‐linear, were employed for the prediction of flow discharge hydrographs at sites receiving significant lateral inflow. The linear model is based on a rating curve and permits a quick estimation of flow at a downstream site. The non‐linear model is based on a multilayer feed‐forward back propagation (FFBP) artificial neural network (ANN) and uses flow‐stage data measured at the upstream and downstream stations. ANN predicted the real‐time storm hydrographs satisfactorily and better than did the linear model. The results of sensitivity analysis indicated that when the lateral inflow contribution to the channel reach was insignificant, ANN, using only the flow‐stage data at the upstream station, satisfactorily predicted the hydrograph at the downstream station. The prediction error of ANN increases exponentially with the difference between the peak discharge used in training and that used in testing. ANN was also employed for flood forecasting and was compared with the modified Muskingum model (MMM). For a 4‐h lead time, MMM forecasts the floods reliably but could not be applied to reaches for lead times greater than the wave travel time. Although ANN and MMM had comparable performances for an 8‐h lead time, ANN is capable of forecasting floods with lead times longer than the wave travel time. Copyright © 2007 John Wiley & Sons, Ltd.     

Aquifer hydraulic conductivity prediction via coupling model of MCMC-ANN

Grain-size distribution data,as a substitute for measuring hydraulic conductivity(K),has often been used to get K value indirectly.With grain-size distribution data of 150 sets of samples being input data,this study combined the Artificial Neural Network technology(ANN)and Markov Chain Monte Carlo method(MCMC),which replaced the Monte Carlo method(MC)of Generalized Likelihood Uncertainty Estimation(GLUE),to establish the GLUE-ANN model for hydraulic conductivity prediction and uncertainty analysis.By means of applying the GLUE-ANN model to a typical piedmont region and central region of North China Plain,and being compared with actually measured values of hydraulic conductivity,the relative error ranges are between 1.55%and 23.53%and between 14.08%and 27.22%respectively,the accuracy of which can meet the requirements of groundwater resources assessment.The global best parameter gained through posterior distribution test indicates that the GLUEANN model,which has satisfying sampling efficiency and optimization capability,is able to reasonably reflect the uncertainty of hydrogeological parameters.Furthermore,the influence of stochastic observation error(SOE)in grain-size analysis upon prediction of hydraulic conductivity was discussed,and it is believed that the influence can not be neglected.     

贵州省乌江流域土壤侵蚀模拟——基于GIS、RUSLE和ANN技术的研究

本研究在GIS技术支撑下选择RUSLE模型作为基础模型,估算乌江流域20世纪80年代和90年代年均土壤侵蚀量,结合ANN技术,预测2001—2010年乌江流域的土壤侵蚀量,分析了该流域近30年来土壤侵蚀动态变化规律,以期为研究区土壤侵蚀防治工作提供理论依据。研究结果表明:应用RUSLE模型计算乌江流域年均土壤侵蚀模数,计算结果和以往土壤侵蚀调查估计的结果比较吻合,但由于RUSLE模型不计算重力侵蚀,因此计算结果仍与实测输沙模数有所出入。90年代潜在土壤侵蚀模数比80年代高,流域潜在土壤侵蚀呈增加趋势,其中三岔河流域和马蹄河/印江河流域年均潜在土壤侵蚀模数最高。3种主要土地覆被类型中,林地的土壤保持量最大,耕地次之,草地最少,这与非喀斯特地区在水土保持效果上通常林地草地旱地的结论有所不同。通过构建BP神经网络,预测得到乌江流域2001—2010年土壤侵蚀模数,结果显示,21世纪前10年,流域土壤侵蚀模数大幅降低,流域年均土壤侵蚀模数由90年代的23.13 t/(hm2·a)降低为1.01 t/(hm2·a)。三岔河流域的水土流失得到了控制,黔西、金沙、息烽、修文、贵阳、平坝、思南、石阡、沿河和松桃等县市应是"十二五"期间的水土流失重点治理对象。     

Neural network model for discharge and water-level prediction for Ramganga River catchment of Ganga Basin,India

ABSTRACT

Discharges and water levels are essential components of river hydrodynamics. In unreachable terrains and ungauged locations, it is quite difficult to measure these parameters due to rugged topography. In the present study an artificial neural network model has been developed for the Ramganga River catchment of the Ganga Basin. The modelled network is trained, validated and tested using daily water flow and level data pertaining to 4 years (2010–2013). The network has been optimized using an enumeration technique and a network topology of 4-10-2 with a learning rate set at 0.06, which was found optimum for predicting discharge and water-level values for the considered river. The mean square error values obtained for discharge and water level for the tested data were found to be 0.046 and 0.012, respectively. Thus, monsoon flow patterns can be estimated with an accuracy of about 93.42%.

Editor M.C. Acreman; Associate editor E. Gargouri     

Precipitation-runoff simulation for a Himalayan River Basin, India using artificial neural network algorithms

In the Himalayan regions, precipitation-runoff relationships are amongst the most complex hydrological phenomena, due to varying topography and basin characteristics. In this study, different artificial neural networks (ANNs) algorithms were used to simulate daily runoff at three discharge measuring sites in the Himalayan Kosi River Basin, India, using various combinations of precipitation-runoff data as input variables. The data used for this study was collected for the monsoon period (June to October) during the years of 2005 to 2009. ANNs were trained using different training algorithms, learning rates, length of data and number of hidden neurons. A comprehensive multi-criteria validation test for precipitation-runoff modeling has been undertaken to evaluate model performance and test its validity for generating scenarios. Global statistics have demonstrated that the multilayer perceptron with three hidden layers (MLP-3) is the best ANN for basin comparisons with other MLP networks and Radial Basis Functions (RBF). Furthermore, non-parametric tests also illustrate that the MLP-3 network is the best network to reproduce the mean and variance of observed runoff. The performance of ANNs was demonstrated for flows during the monsoon season, having different soil moisture conditions during period from June to October.     

ANN技术在地下水含水量预测建模中的研究与应用

将人工神经网络(ANN)技术引入到地下水含水量预测工作,以华北平原和河套平原为试验场,以若干已知钻孔为验证,采用激电和电阻率测深等地面物探方法获取视电阻率ρS、视极化率ηS、半衰时Th、衰减度D和偏离度σ等参数为输入神经元对单孔单位涌水量建立人工神经网络预测模型。同时,为消除不同地区矿化度的影响,通过实验对比引入综合参数T",改良了输入神经元的配比。最终建立以半衰时Th、衰减度D、偏离度σ和综合参数T"为输入神经元的含水量预测模型,进一步提高了预测精度。通过检验,发现所建立的模型对平原地区进行含水量的定量预测有着较好的效果,为含水量预测工作研究与发展带来了新理念、打开了新思路。     

��BP�����缼��̽���봨����ǰ�����NmF2�쳣�Ŷ�

????UCAR??????????F2???????????????NmF2??????????????缼???????????????NmF2???????????DOY???????LT??????LON??γ??LAT??F10.7????????FLUX???????????NmF2???????????????????NmF2????????ο????????????????????????????2008??5??12????7.9??????????и??????NmF2???????6??4??6??8?????С?30%???????3??2??9??10????????????40%??     

Replication of a terrain stability mapping using an Artificial Neural Network

Subjective geomorphic mapping is a method commonly used for landslide hazard zonation. This method relies heavily on the skills and experience of the mapper, and therefore, its major drawbacks are the high costs and lack of consistency between products generated by different terrain mappers. In this study a method for cost-effective and consistent replication of subjective geomorphic mappings is demonstrated, by using a type of Artificial Neural Network named Learning Vector Quantization. This paper presents a study conducted in the Canadian province of British Columbia employing a high-quality data set. By utilizing Learning Vector Quantization, stable and unstable terrains were delineated with a similarity of approximately 91%, compared to the mapping produced by terrain specialists. Also, in this process, slope, elevation, aspect, and existing geomorphic processes were identified as the terrain attributes that contributed most to the quality of the mapping.     

Effects of sample size on the accuracy of geomorphological models

Commonly, the most costly part of geomorphological distribution modelling studies is gathering the data. Thus, guidance for researchers concerning the quantity of field data needed would be extremely practical. This paper scrutinises the relationship between the sample size (the number of observations varied from 20 to 600) and the predictive ability of the generalized linear model (GLM), generalized additive model (GAM), generalized boosting method (GBM) and artificial neural network (ANN) in two data settings, i.e., independent and split-sample approaches. The study was performed using empirical data of periglacial processes from an area of 600 km² in northernmost Finland at grid resolutions of 1 ha (100 × 100 m) and 25 ha (500 × 500 m). A rather sharp increase in the predictive ability of the models was observed when the number of observations increased from 20 to 100, and the level of robust predictions was reached with 200 observations. The result indicates that no more than a few hundred observations are needed in geomorphological distribution modelling at a medium scale resolution (ca. 0.01–1 km²).