基于DEM的黑龙江宾州河流域水系提取试验研究

基于数字高程模型(digital elevation models,DEM)提取河网及相关流域信息是数字水文研究中的热点问题。以东北黑土区宾州河流域为研究区域,采用5m分辨率的DEM作为研究数据,运用ArcGIS9.2中的Hydrology水文处理工具包(Arc Hydro Tools)提取研究区河网特征。研究结果表明,分别采用0.75,1.25,2,2.5,3.75和5km2的集水面积阈值提取流域河网特征时,不同集水面积阈值对确定主河道的空间位置影响较小,但提取的数字流域河网特征会发生较大变化。基于河网密度与集水面积阈值及河源密度与集水面积阈值的关系探讨了适宜的集水面积阈值确定方法,发现取集水面积阈值为2km2时提取的河网特征与实际水系特征基本吻合,精度达95%以上。因此,基于Arc Hydro Tools的流域特征自动提取是切实可行的。     

Estimating and modeling spatio-temporal correlation structures for river monitoring networks

The European environmental legislation forces local authorities to improve the river water quality. In order to assess the presence of trends or the effect of certain actions on the river water quality, a statistical methodology is needed which can deal with data originating from river monitoring networks. Since both temporal and spatial components affect the output of such a monitoring network, their dependence structure has to be modeled. Current spatio-temporal models used for the analysis of data arising from environmental studies are not appropriate because they do not deal properly with the particular spatial dependence structure underlying river monitoring networks. In this article, a state-space model is developed in which the state variable is defined by a directed acyclic graph (DAG) derived from the river network topology. In reality the dependence structure based on the DAG may be obscured by environmental factors. This is taken into account by embedding the state variable in an observation model. Finally, the state-space model is extended with a linear model for the mean. An efficient ECM-like algorithm is proposed for parameter estimation, using the Kalman filter and smoother in both E- and CM-steps. In a case-study the method is applied to assess the effect of the activation of a waste water treatment plant on the dissolved oxygen concentration, in time as well as in space.     

Deriving river networks and catchments at the European scale from medium resolution digital elevation data

The extraction of drainage networks and catchment boundaries from digital elevation models (DEMs) has received considerable attention in recent years and is recognized as a viable alternative to traditional surveys and the manual evaluation of topographic maps. However, most studies have covered limited areas due to the lack of detailed information and/or the lack of highly efficient algorithms. In this paper we present an application that delineates river networks and catchment boundaries across the European continent from a medium resolution (250 m) DEM. We exploit novel algorithms based on the concepts of mathematical morphology and implement a landscape stratification for drainage density.A flow direction grid is computed using an efficient algorithm for the removal of spurious pits. River networks are then derived by imposing a variable threshold for the minimum contributing area needed to form and maintain a channel. This is achieved through a landscape stratification that reflects the ability of the terrain to develop different drainage densities. It is shown that the analysis of environmental characteristics coupled with the analysis of local slope versus contributing area enables river network mapping with a spatially varying drainage density. The result has been validated by comparing the derived data to digital river and catchment data from other sources and with varying scales of observation.     

融合多尺度特征的冬小麦空间分布提取方法

如何获取到高质量的特征是从遥感影像中提取高精度的农作物空间分布的关键，该研究针对如何利用哨兵2A（Sentinel-2A）影像提取高精度的冬小麦空间分布开展研究。针对影像中存在的数据空间尺度不一致的问题，以生成式对抗网络为基础建立了降尺度模型REDS（Red Edge Down Scale），用于将B5、B6、B7、B11 4个通道的空间分辨率从20 m降为10 m；然后利用卷积神经网络构建了逐像素分割模型REVINet（Red Edge and Vegetation Index Feature Network），REVINet以10m分辨率的B2、B3、B4、B5、B6、B7、B8、B11，以及提取出的增强植被指数、归一化植被指数和归一化差值红边指数组合作为输入，进行逐像素分类。选择ERFNet、U-Net和RefineNet作为对比模型同REVINet开展对比试验，试验结果表明，该研究提出的方法在召回率（92.15%）、精确率（93.74%）、准确率（93.09%）和F1分数（92.94%）上均优于对比方法，表明了该研究在从Sentinel-2A中提取冬小麦空间分布方面具有明显的优势。     

Holocene sediment budgets for upland catchments: The problem of soilscape model and data availability

This study presents a Holocene sediment budget for the upland catchment of the Speyerbach in the Palatinate Forest, southwestern Germany. The influences of both, data availability and the choice of an appropriate soilscape model on the sediment budget calculations are investigated. For budget calculations the spatial distribution of soils was derived from the soil map 1:50,000. Thickness values of soil truncation and colluvial burial were extracted from two soil data sets with varying information content. Data processing contained a disaggregation of the soil map with the help of land use data. In order to model the sediment budget a reference soil thickness (a so-called soilscape model), which represents the initial conditions is necessary. As upland soils are developed in periglacial solifluction sheets, the main solifluction sheet (MSS, “Hauptlage”) showing a constant thickness and being affected by soil forming processes was chosen.     

通量贡献区叶面积指数空间分布的测定

使用循环采样设计方案,对海北沼泽化湿地生态系统涡度相关通量观测塔的通量贡献区内叶面积指数进行了实地调查。根据样方叶面积指数的实地测量值和样方的GPS空间定位信息,利用空间插值方法绘制了通量贡献区内的叶面积指数空间分布图。并基于数字摄影与地理信息系统技术,提出了一种测量速度快、计算精度高、适合多种植物叶片叶面积测量的新方法。该方法在ArcGIS的GRID模块下对叶片图像进行格式转化和重新采样处理,使用自行编写的色阶诊断程序提取图像中叶片的叶面积。研究结果表明:此种方法叶面积测量结果与LI-3000A叶面积仪所测的结果具有很好的吻合性,两者的线性回归方程决定系数R2为0.98,叶面积的测量精度完全可以达到实际应用中的测量要求。     

基于轻量和积网络及无人机遥感图像的大豆田杂草识别

为提高机器视觉在无人机等小型嵌入式设备中杂草识别的准确率,该文以大豆苗中常见禾本科杂草和阔叶型杂草为研究对象,针对传统和积网络在图像分类任务中模型参数多、训练时间长、含有较多冗余节点和子树的问题,该文改进传统和积网络的学习过程,提出一种以小批量数据作为输入的轻量和积网络。在结构学习中,当积节点作用域内的变量个数小于一定阈值时,合并积节点为多元叶节点,否则将积节点重组为和积混合结构,并对边缘节点进行裁剪,有效降低了模型的参数量和复杂度。在参数学习中,提出贝叶斯矩匹配更新网络参数,使得模型对小样本的学习效率更高。最后结合K均值聚类算法应用于无人机图像中的杂草识别。试验结果表明,利用该方法对无人机图像中大豆苗、禾本科杂草、阔叶型杂草以及土壤的平均识别准确率达99.5%,高于传统和积网络和传统AlexNet。并且模型平均参数量仅为传统和积网络的33%,内存需求最大时减少了549 M,训练时间最多减少了688.79 s。该研究可为轻量和积网络模型在无人机喷洒农药中的杂草识别提供参考。     

基于轮廓不同的DEM对宁波市姚江流域平原河网的提取研究

[目的]针对目前基于DEM提取平原河网所遇到的困境,采用新的视角提取平原河网,为完善水系自动化提取方法研究提供依据。[方法]采用Acr Hydro Tools和SWAT工具在轮廓不同的DEM下,对宁波姚江流域进行河网提取和流域确定。再以提取后的流域总面积与河流总长度为定量分析的主要指标,用相对误差公式、河网套合差公式和Visual Similarity Duplicate Image Finder软件进行综合评价与误差分析。[结果]在真实河网辅助下,Acr Hydro Tools与SWAT均是在规整型DEM基础上的提取效果很差。在临近型DEM上所提取的河网、流域和流域总出口与实际相符且精度、相似度都很高。[结论]用Acr Hydro Tools和SWAT两种工具提取流域水系都是可取的,但SWAT更加便捷;不同轮廓的DEM以及其是否与流域总出口相切会极大地影响河网提取的质量。     

Testing Backpropagation Neural Network Approach in Interpolating Missing Daily Precipitation

This study proposes a method, backpropagation (BP) neural network, for interpolating missing values in daily precipitation time series. Firstly, the BP neural network is adopted to interpolate missing daily rainfall data at three selected stations in Yantai, Shandong, China. Then, the temporal and spatial variations in precipitation extremes across Shandong are analyzed by utilizing the complete daily rainfall dataset derived from accurate propagation at 24 meteorological stations. The results show that the long-term trends in five selected extreme precipitation indices calculated from interpolated daily rainfall data are generally consistent with those from original nonmissing values. And the spatial patterns of trends in precipitation extremes also show better performance for BP neural network approach in interpolating missing daily rainfall gaps. Those suggest that this BP neural network algorithm can obtain a good fit in terms of space-time variability of regional precipitation extremes, in case that the correlation coefficients between the target stations with missing values and reference stations with complete daily rainfall dataset are relatively large. These findings could be crucial for investigating regional frequency of heavy rainfall and water resource management.     

DEM水平分辨率对流域特征提取的影响分析

在基于数字高程模型（DEM）的流域地形分析中,栅格DEM的分辨率对分析结果具有很大的影响。以舒城县内面积为7.9 km²的龙潭小流域作为研究区域,分别对分辨率为5 m×;5 m～50 m×;50 m的10幅DEM数据进行流域特征提取和分析,计算地形指数并分析其分布特征。研究表明:在小流域尺度上,DEM分辨率的变化对流域面积、最长河道长度等参数影响不大,对河道总长度、河网密度及平均坡度等参数有较大影响,对流域特征的提取影响规律为DEM分辨率越低,提取的流域特征越粗糙;且分辨率低于15 m的DEM计算出的地形指数值较为离散、概率分布曲线形状较之5 m分辨率DEM的地形指数概率分布曲线有明显不同,表明分辨率低于15 m的DEM对小流域水文地形信息的反映较差,为小流域尺度水文模型DEM数据适宜分辨率选取提供一定的理论依据。     

基于SWAT模型的平原河网区水文过程分布式模拟

目前普遍使用的基于SWAT(soil and water assessment tool)模型的分布式建模方法仅适用于山地、丘陵等高差较大的地区,对于具有复杂水文结构特点的平原河网区尚无有效的解决方案。该文选择太湖流域湖西区作为研究区,基于SWAT模型探索出一套完整的针对平原河网区的分布式建模方案。该方案采用概化、打断的方式将交叉、环状河网处理成单一的枝状河网,采用按河道分流比例跨子流域调水的方式来还原原始河道的传输过程,采用添加"虚拟水库"的方式来模拟人工圩区内外的水量交换。通过对太湖湖西区2008-2010年的月均径流量进行模拟验证,根据模拟值和观测值计算的荣登桥、胡家圩及宜兴3个水文站的相关系数r和纳什(Nash-Sutcliffe)系数Ens分别为0.94、0.95、0.93和0.84、0.80、0.67,说明了这种建模方法在平原河网区具有较好的适用性。     

Spatial prediction of soil classes using digital terrain analysis and multinomial logistic regression modeling integrated in GIS: Examples from Vestfold County,Norway

The main objectives of this study were to model the relationship between WRB-1998 soil groups and terrain attributes and predict the spatial distribution of the soil groups using digital terrain analysis and multinomial logistic regression integrated in GIS in the Vestfold County of south-eastern Norway. A digital elevation model of 25 meter grid resolution was used to derive fifteen terrain attributes. A digitized soil map of thirteen WRB soil groups at the scale of 1:25,000 was used to obtain the reference soil data for model building and validation. First, the relationships between the soil groups and the terrain attributes were modeled using multinomial logistic regression. Then, the probability that a given soil type is present at a given pixel was determined from the logit models in ARCGIS to continuously map each soil group's spatial distribution. Elevation, flow length, duration of daily direct solar radiation, slope, aspect and topographic wetness index were found to be the most significant terrain attributes correlating with the spatial distribution of the soil groups. The prediction showed higher mean probability values for each soil group in the areas actually covered by that soil group compared to other areas, indicating the reliability of the prediction. However, the prediction performed poorly for soil groups that are not greatly influenced by topography but by other factors such as human activities.     

Spatial prediction on a river network

This article develops methods for spatially predicting daily change of dissolved oxygen (Dochange) at both sampled locations (134 freshwater sites in 2002 and 2003) and other locations of interest throughout a river network in South East Queensland, Australia. In order to deal with the relative sparseness of the monitoring locations in comparison to the number of locations where one might want to make predictions, we make a classification of the river and stream locations. We then implement optimal spatial prediction (ordinary and constrained kriging) from geostatistics. Because of their directed-tree structure, rivers and streams offer special challenges. A complete approach to spatial prediction on a river network is given, with special attention paid to environmental exceedances. The methodology is used to produce a map of Dochange predictions for 2003. Dochange is one of the variables measured as part of the Ecosystem Health Monitoring Program conducted within the Moreton Bay Waterways and Catchments Partnership.     

露天苗圃灌溉控制系统网关节点调度模型

为了提高苗圃灌溉控制系统的稳定性和健壮性,该文在无线传感器网络中,以网关节点调度为目标,提出了一种新的基于马尔可夫链的网关节点调度方法。首先,理论上分析单个网络节点的存在状态,及其接入行为;其次,以网关和终端节点共同状态为基础,通过单个节点的状态转移概率关系,得到系统平衡方程;第三,根据此平衡方程进行节点的调度,以及描述系统工作过程;最后对得出的系统调度模型进行性能参数估计。试验建立一个无线传感器网络仿真平台,对该模型进行从带宽利用率、被迫中断概率、以及网络容量和阻塞率进行比较分析。结果表明随着用户的呼叫率增多,带宽的利用率、终端被迫中断概率和系统容量都随之增长,而当呼叫量增长到一定程度时,系统容量会慢慢趋于平缓。该模型在提高传感器网络带宽利用率的同时保证了一定的系统中断概率。     

黄河粗泥沙集中来源区区域地形因子比较研究

在黄河粗泥沙集中来源区的清水川流域,以1∶10 000比例尺地形图上的小流域为基本单元,在1∶250 000比例尺DEM上提取地表起伏度、高程变异系数和等高线密度等区域土壤侵蚀地形因子,以1∶10 000比例尺计算的地形因子为真值,拟合各区域土壤侵蚀地形因子的计算方法。结果表明,等高线密度是合适的区域土壤侵蚀地形因子计算指标。应用该方法在皇甫川下游进行检验,模拟结果误差小于10%,效果良好,可作为土壤侵蚀区域地形因子计算时指标选择的参考。     

小流域尺度土壤可蚀性(K值)的变异及不同采样密度对其估值精度的影响

以观音寺小流域为例,通过高密度采样,利用EPIC模型中土壤可蚀性K值计算公式计算了可蚀性K值,并运用逆距离加权插值方法生成了研究区的K值分布图。经典统计学方法分析表明,在小流域尺度下土壤可蚀性K值存在很强的空间变异性,其变异系数达39.94%。因此,定量监测、评价流域水土流失时,K值的这种空间变异性不能忽略。对不同采样密度下K值空间估值精度的评价表明,当采样点数目均匀地减少一半时,即4个/km2的样点密度下,K值估计误差仍然可以控制在25%以内。     

Digital soil mapping using artificial neural networks

In the context of a growing demand of high‐resolution spatial soil information for environmental planning and modeling, fast and accurate prediction methods are needed to provide high‐quality digital soil maps. Thus, this study focuses on the development of a methodology based on artificial neural networks (ANN) that is able to spatially predict soil units. Within a test area in Rhineland‐Palatinate (Germany), covering an area of about 600 km², a digital soil map was predicted. Based on feed‐forward ANN with the resilient backpropagation learning algorithm, the optimal network topology was determined with one hidden layer and 15 to 30 cells depending on the soil unit to be predicted. To describe the occurrence of a soil unit and to train the ANN, 69 different terrain attributes, 53 geologic‐petrographic units, and 3 types of land use were extracted from existing maps and databases. 80% of the predicted soil units (n = 33) showed training errors (mean square error) of the ANN below 0.1, 43% were even below 0.05. Validation returned a mean accuracy of over 92% for the trained network outputs. Altogether, the presented methodology based on ANN and an extended digital terrain‐analysis approach is time‐saving and cost effective and provides remarkable results.     

基于人工神经网络的田间信息插值方法研究

提出了一种基于人工神经网络的田间信息插值新方法，并利用ArcView3.2软件绘制碱解氮的BP神经网络插值空间分布图和球状插值分布图，并对BP神经网络插值方法和克立格球状插值方法的结果进行了误差分析。结果表明，BP神经网络的插值方法优于克立格球状插值法，该方法有利于田间信息空间分布特性准确、直观的表达，有利于农田精确施肥、灌溉、播种等精细农业生产管理。     

Integrating soil map information in modelling the spatial variation of continuous soil properties

This paper presents two indicator algorithms that integrate soil map information into modelling the spatial variation of continuous soil properties: these are simple indicator kriging with varying means and the Markov–Bayes algorithm. Both methods are used to evaluate probabilities for copper and cobalt deficiencies in the Borders Region of Scotland. Results are compared with maps obtained by the polygonal method (Thiessen polygons) and an indicator kriging algorithm that does not use soil map information. Accounting for soil map information is shown to improve delineation of the deficient areas, especially where the sampling is sparse. Test locations are classified as deficient or not so as to minimize an expected cost of mis-classification that is derived from local probability distributions of copper or cobalt and functions measuring the cost of overestimating or underestimating metal concentrations. The comparison of classification results with actual copper and cobalt concentrations at test locations shows that the two proposed algorithms can decrease substantially the economic loss attached to misclassification.     

LiDAR-derived slopes for headwater channel network analysis

Slope is one of the most important distinguishing features for channel morphology. Variations in the computation of slope from a digital elevation model can affect a wide range of hydrogeomorphically derived applications. We compare different methods for computing channel slope using LiDAR-derived digital terrain models (DTMs) with varying resolutions. We chose a headwater basin of the Eastern Italian Alps, characterized by a dense ephemeral colluvial network and a main alluvial channel as our study area. The identified alluvial morphologies are characteristic of steep mountain streams, namely, cascades and step pools. Field surveys were carried out along the main channel and in some small tributaries. Results indicate that a single method for slope calculation cannot estimate channel slope at the hydrographic network scale. The differential geometry approach for slope calculation tends to overestimate field-surveyed channel slope values for all the DTM resolutions (1, 2, 5 m). When a trigonometric approach for slope calculation is applied, 2 and 5 m DTM resolutions give more consistent results. Nevertheless, a reliable channel slope can be derived from a DTM with an appropriate resolution by choosing a suitable method only after considering the channel width.