Neural network ensemble residual kriging application for spatial variability of soil properties

High quality, agricultural nutrient distribution maps are necessary for precision management, but depend on initial soil sample analyses and interpolation techniques. To examine the methodologies for and explore the capability of interpolating soil properties based on neural network ensemble residual kriging, a silage field at Hayes, Northern Ireland, UK, was selected for this study with all samples being split into independent training and validation data sets. The training data set, comprised of five soil properties: soil pH, soil available P, soil available K, soil available Mg and soil available S,was modeled for spatial variability using 1) neural network ensemble residual kriging, 2) neural network ensemble and 3) kriging with their accuracies being estimated by means of the validation data sets. Ordinary kriging of the residuals provided accurate local estimates, while final estimates were produced as a sum of the artificial neural network (ANN) ensemble estimates and the ordinary kriging estimates of the residuals. Compared to kriging and neural network ensemble,the neural network ensemble residual kriging achieved better or similar accuracy for predicting and estimating contour maps. Thus, the results demonstrated that ANN ensemble residual kriging was an efficient alternative to the conventional geo-statistical models that were usually used for interpolation of a data set in the soil science area.     

基于GIS技术的红壤资源划分与评估

A small scale red soil resources information system(RSRIS) with applied mathematical models was developed and applied in red soil resources(RSR) classification and evaluation,taking Zhejiang Province,a typical distribution area of red soil,as the study area.Computer-aided overlay was conductied to classifty RSR types.The evaluation was carried out by using three methods,i.e.,index summation,square root of index multiplication and fuzzy comprehensive assessment,with almost identical results,The result of index summation could represent the basic qualitatie condition of RSR,that of square root of index miltiplication reflected the real condition of RSR qualitative rank,while fuzzy comprehensive assessment could satisfactorily handle the relationship between the evaluation factors and the qualitative rank of RSR,and therefore it is a feasible method for RSR evaluation.     

基于GIS和改进型土地生态适宜性评价模型的茶叶适宜性综合评价

Tea(Camellia sinensis) is one of the most valuable cash crops in southern China;however,the planting distribution of tea crops is not optimal and the production and cultivation regions of tea crops are restricted by law and custom.In order to evaluate the suitability of tea crops in Zhejiang Province,the annual mean temperature,the annual accumulated temperature above 10 C,the frequency of extremely low temperature below 13 C,the mean humidity from April to October,slope,aspect,altitude,soil type,and soil texture were selected from climate,topography,and soil factors as factors for land ecological evaluation by the Delphi method based on the ecological characteristics of tea crops.These nine factors were quantitatively analyzed using a geographic information system(GIS).The grey relational analysis(GRA) was combined with the analytic hierarchy process(AHP) to address the uncertainties during the process of evaluating the traditional land ecological suitability,and a modified land ecological suitability evaluation(LESE) model was built.Based on the land-use map of Zhejiang Province,the regions that were completely unsuitable for tea cultivation in the province were eliminated and then the spatial distribution of the ecological suitability of tea crops was generated using the modified LESE model and GIS.The results demonstrated that the highly,moderately,and non-suitable regions for the cultivation of tea crops in Zhejiang Province were 27 552.66,42 724.64,and 26 507.97 km 2,and accounted for 28.47%,44.14%,and 27.39% of the total evaluation area,respectively.Validation of the method showed a high degree of coincidence with the current planting distribution of tea crops in Zhejiang Province.The modified LESE model combined with GIS could be useful in quickly and accurately evaluating the land ecological suitability of tea crops,providing a scientific basis for the rational distribution of tea crops and acting as a reference to land policy makers and land use planners.     

免疫算法在土壤资源质量评价中的应用

Based on the geographic information system （GIS） technology, ArcInfo software was adopted to collect, process and analyze spatial data of Guangdong Province for an evaluation of soil resource quality. The overlay analysis method was used in combining evaluation factors of Guangdong soil resource quality to determine the evaluation units. Because of its favorable convergent speed and its ability to search solutions, the immune algorithm was applied to the soil resource quality evaluation model. At the same time, the evaluation results of this newly proposed method were compared to two other methods： sum of index and fuzzy synthetic. The results indicated that the immune algorithm reflected the actual condition of soil resource quality more exactly.     

结合统计和数字地形数据的可视化方法预测土壤深度

Information about the spatial distribution of soil attributes is indispensable for many land resource management applications; however, the ability of soil maps to supply such information for modern modeling tools is questionable. The objectives of this study were to investigate the possibility of predicting soil depth using some terrain attributes derived from digital elevation models (DEMs) with geographic information systems (GIS) and to suggest an approach to predict other soil attributes. Soil depth was determined at 652 field observations over the Al-Muwaqqar Watershed (70 km²) in Jordan. Terrain attributes derived from 30-m resolution DEMs were utilized to predict soil depth. The results indicated that the use of multiple linear regression models within small watershed subdivisions enabled the prediction of soil depth with a difference of 50 cm for 77% of the field observations. The spatial distribution of the predicted soil depth was visually coincided and had good correlations with the spatial distribution of the classes amalgamating three terrain attributes, slope steepness, slope shape, and compound topographic index. These suggested that the modeling of soil-landscape relationships within small watershed subdivisions using the three terrain attributes was a promising approach to predict other soil attributes.     

自动土壤图基于知识的分类

A machine-learning approach was developed for automated building of knowledge bases for soil resources mapping by using a classification tree to generate knowledge from training data. With this method, building a knowledge base for automated soil mapping was easier than using the conventional knowledge acquisition approach. The knowledge base built by classification tree was used by the knowledge classifier to perform the soil type classification of Longyou County, Zhejiang Province, China using Landsat TM bi-temporal images and GIS data. To evaluate the performance of the resultant knowledge bases, the classification results were compared to existing soil map based on a field survey. The accuracy assessment mad maalysis of the resultant soil maps suggested that the knowledge bases built by the machine-learning method was of good quality for mapping distribution model of soll classes over the study area.     

基于模糊逻辑的北京城市边缘区土壤重金属污染空间预测

Fuzzy classification combined with spatial prediction was used to assess the state of soil pollution in the peri-urban Beijing area. Total concentrations of As, Cr, Cd, Hg, and Pb were determined in 220 topsoil samples （0-20 cm） collected using a grid design in a study area of 2 600 kin2. Heavy metal concentrations were grouped into three classes according to the optimum number of classes and fuzziness exponent using the fuzzy comean （FCM） algorithm. Membership values were interpolated using ordinary kriging. The polluted soils of the study area induced by the measured heavy metals were concentrated in the northwest corner and eastern part, especially the southeastern part close to the urban zone, whereas the soils free of pollution were mainly distributed in the southwestern part. The soils with potential risk of heavy metal pollution were located in isolated spots mainly in the northern part and southeastern corner of the study region. The FCM algorithm combined with geostatistical techniques, as compared to conventional single geostatistical kriging methods, could produce a prediction with a quantitative uncertainty evaluation and higher reliability. Successful prediction of soil pollution achieved with FCM algorithm in this study indicated that fuzzy set theory had great potential for use in other areas of soil science.     

湿地土壤质量退化的模糊综合评价

Wetland soil quality degradation caused by large-scale agricultural reclamation on the Sanjiang Plain of Northeast China has been widely reported. A relative soil quality evaluation (RSQE) model and a projection pursuit evaluation (PPE) model based on real-coded accelerating genetic algorithm were introduced to evaluate quality variations in top layers of the main wetland soils on the Sanjiang Plain in 1999 and 2003, respectively. As soil quality degradation boundaries were vague, this study established two fuzzy synthetic evaluation (FSE) models based on the original data and criteria used in the RSQE and PPE models. The outputs of the two FSE models were obtained by choosing two fuzzy composite operators M (∧,∨) and M (?,⊕). Statistical analysis showed that the results of the FSE, RSQE, and PPE models were correlated. In particular, outputs of the FSE model using M (?,⊕) were significantly correlated with those of the RSQE model with r = 0.989 at P < 0.01. Compared with RSQE and PPE models, the FSE model may be more objective in showing soil quality variations and was closer to the natural situation, showing the feasibility and applicability of the FSE model in evaluating soil quality degradation. However, the choice of composite operator was of critical importance. The study of wetland soil quality degradation on the Sanjiang Plain was of scientific and practical significance for protection and management of soils and for sustainable development of agriculture in this area in the future.     

土地质量评估与监测: 土壤科学面临的新挑战

Sustainable land management (SLM)is the key to harmonizing environmetal and ecolgical concerns of society with the economic realities of producing adequate food and fiber of high quality and ensuring a absi minimal quality of life.The aim of SLM is to maintain the integrity of the biophysical land resource base,but it can only be realized if land users understand the impacts of land management options on their lands but also on other off-site areas and can optimize the socioeconomic and environmetal benefits of their choice.To Facilitate this,the Contribution of soil suvery organizations would be through the assessment and monitoring of land quality.Land quality is a measure of the ability of land to perfor specific functions and is derived by an integration of soil survey information with other environmental,and if necessary,socioeconomic information.The desired reliability influences the operational scale of the assessment,Such an assessment would assist in:1) locating homologous areas for research sites or for transferring technologies;2) providing the geographic basis for systems analysis(e.g.modeling);3) serving as a basis for local,natinal and global resource assessment and monitoring;4) providing an ecosystem context for land use,assessments of temporal and spatial variability,and impact of human interventions;5)serving as a framework for more detailed assessment for all levels of interest;and 6) evaluating global issues such as food security,impacts of climate change,biodiversity montoring,and addressing desertification.Based on an evaluation of the progress made in soil resource inventories and considering the demands of the environment focused world,the paper considers the need for counteries to mount such a program.The authors believe that this is the next demand of soil science and that we can fulfill our soical contract by periodically providing such information on the state of a nation‘s land resource.     

基于SFAM神经网络集成的土地评价

SFAM(Simplified Fuzzy ARTMAP,简化的模糊ARTMAP)神经网络具有自组织反馈、增量式学习和高度复杂映射等特点,是一种较BP神经网络和RBF神经网络等前馈神经网络更优秀的自组织神经网络.为克服SFAM神经网络受输入样本顺序的影响,提高土地评价的精度,提出利用SFAM神经网络集成进行土地评价的方法.并用SFAM神经网络、SFAM神经网络集成、BP神经网络、BP神经网络集成、RBF神经网络和RBF神经网络集成等方法对广东省中山市的土地进行了评价,对评价结果进行了分析和比较,结果表明SFAM神经网络具有比BP神经网络和RBF神经网络更优越的评价性能;对于这三种不同的神经网络,神经网络集成的土壤评价精度分别高于单个神经网络的精度.     

从神经网络中抽取土地评价模糊规则

为了明确土地评价中所训练神经网络的含义，使土地评价工作者可轻松地理解、判断所得到土地评价模型的正确性和合理性，提出从神经网络中抽取土地评价模糊规则的方法。现有的大多数从神经网络中提取方法，神经网络的输入属性要么局限于连续的，要么只适应于离散的，而土地评价因子往往既包含连续的又包含离散的、标称的，该文首先提出了一种输入属性值适应于这三种类型数据的模糊神经网络建立方法，进而给出一种从建立的神经网络中抽取其中较主要模糊规则的算法。试验表明，所提出的土地评价方法，可直接从样本中学习评价规律，使土地评价工作者易于理解，当出现抽取的规则与实际情况不吻合时，可重新训练神经网络和抽取规则，所得到的评价结果比BP网络的评价结果更准确，从而提高了土地评价的准确性。     

广州市建设用地集约利用评价研究

城市土地的集约利用直接关系到区域土地的可持续利用,如何衡量土地的集约利用程度是一个复杂而重要的过程。以广州市1990—2007年建设用地数据、各类社会经济数据为基础,构建适合广州市建设用地集约利用评价的指标体系,综合运用主成分分析和BP神经网络建立建设用地集约利用评价模型,计算其集约利用程度,并与前人通过模糊综合评价法得出的结论进行比较。结果表明:1990—2007年广州市整体集约利用水平呈上升趋势,1990—1994年、1996—1997年处于粗放利用,1995年、1998—2003年处于中度集约,2004—2007年处于高度集约。主成分分析和BP神经网络的评价方法与模糊综合评价法的结果平均误差为4.37%。     

基于因子分析的Hopfield神经网络在水质评价的应用

针对Hopfield神经网络的过度拟合问题,在因子分析的基础上,结合Hopfield神经网络模型提出了因子分析—Hopfield神经网络模型。以东辽河为例,采用因子分析法确定7个水质评价因子,再建立5×7的Hopfield神经网络进行水质综合评价,并与单一的Hopfield网络和传统的内梅罗指数法的结果进行了比较。结果表明,因子分析—Hopfield神经网络明显优于单一的Hopfield神经网络,不仅在一定程度上弥补了因子分析在实际应用中没有实现水质分级的缺陷,而且有效地降低了Hopfield神经网络的过度拟和的程度,评价结果更为科学合理,为水质综合评价提供了一种新的方法,具有极好的应用前景。     

BP神经网络组合模型在次洪量预测中的应用

[目的]探讨BP神经网络组合模型在次洪量预测中的应用,为黄土高原淤地坝群的安全度汛提供决策依据。[方法]构建基于多元线性回归模型(MLR)和去趋势互相关分析法(DCCA)的BP神经网络组合模型;选择均方差(MSE)、平均绝对误差(MAE)、平均绝对百分比误差(MAPE)以及确定性系数(DC)作为评价指标,与单一模型(多元线性回归模型、BP神经网络模型以及去趋势互相关分析法)进行比较。[结果]BP神经网络组合模型的4项指标MSE,MAE,MAPE和DC分别为2.144,5.453,0.074和0.988,均优于单一模型;模型预测效果从优到劣分别为BP神经网络组合模型、BP神经网络模型、多元线性回归模型和去趋势互相关分析法。[结论]BP神经网络组合模型较单一模型平稳性增强,提高了预测效果,可用于淤地坝群的次暴雨洪量预测。     

黑龙江省黑土区拉林河流域土壤侵蚀强度评价方法比较

为了保护水土资源、改善生态环境,进行区域土壤侵蚀强度评价,以黑龙江省黑土区拉林河流域为研究区,选取坡度、坡向、土壤类型、土地利用状况和标准化植被指数等5项评价指标,分别采用逻辑回归和广义回归神经网络模型,在ArcGIS平台上进行土壤侵蚀强度评价。应用受试者工作特征曲线对2种方法的评价结果进行对比。结果表明:逻辑回归模型和广义回归神经网络模型的受试者工作特征曲线下面积值分别为0.857和0.881,与实际的土壤侵蚀强度情况基本吻合;2种模型的评价结果可以相互校验,广义回归神经网络模型评价结果的精度较高。     

基于神经网络的土壤肥力综合评价

基于BP神经网络算法的模糊综合评判法对深州市土壤肥力进行等级评价,结果表明,该方法比主成分分析改进的模糊综合评判法更客观、稳定,适用性强,深州市至少89.1%的耕地肥力处于中下等水平;同时利用GIS的插值技术,生成土壤肥力评价图,为深州市的土壤改良和测土培肥提供一定依据。     

Spatial prediction of soil nutrient in a hilly area using artificial neural network model combined with kriging

It is widely recognized that using correlated environmental factors as auxiliary variables can improve the prediction accuracy of soil properties. In this study, a radial basis function neural network (RBFNN) model combined with ordinary kriging (OK) was proposed to predict spatial distribution of four soil nutrients based on the same framework used by regression kriging (RK). In RBFNN_OK, RBFNN model was used to explain the spatial variability caused by the selected auxiliary factors, while OK was used to express the spatial autocorrelation in RBFNN prediction residuals. The results showed that both RBFNN_OK and RK presented prediction maps with more details. However, RK does not always obtain mean errors (MEs) which were closer to 0 and lower root mean square errors (RMSEs) and mean relative errors (MREs) than OK. Conversely, MREs of RBFNN_OK were much closer to 0 and its RMSEs and MREs were relatively lower than OK and RK. The results suggest that RBFNN_OK is a more unbiased method with more stable prediction performance as well as improvement of prediction accuracy, which also indicates that artificial neural network model is more appropriate than regression model to capture relationships between soil variables and environmental factors. Therefore, RBFNN_OK may provide a useful framework for predicting soil properties.     

应用集成BP神经网络进行田间土壤空间变异研究

以英国北爱尔兰Hayes的一块牧草地为研究区，将所有样点分为独立的训练和检验数据集，并在训练样点集的基础上设计了其他4种样点布局方案，以研究神经网络集成技术应用于田间土壤性质空间变异性的可能性。与广泛应用的克里格法的试验结果相比，集成BP神经网络的插值结果精度与之基本相当，尤其是在样点分布较稀疏和样点数较少的情况下，集成BP网络表现出明显的优势；由于神经网络集成方法对样本数据的分布没有任何要求，因此具有较广泛的应用前景和潜力，并在不符合克里格法对样本数据分布要求的情况下是一种可行的替代方法。     

基于改进遗传算法的温湿度模糊神经网络控制器

为了创造适合作物生长的环境,针对温室系统的特点,该文提出了一种基于改进遗传算法的模糊神经网络控制器,利用改进遗传算法训练模糊神经网络模型,采用此模糊神经网络控制器控制温室系统,由数值实验可以看到采用此控制器的温室系统具有响应速度快、过程平稳、编程简单的特点.