Gully erosion spatial modelling: Role of machine learning algorithms in selection of the best controlling factors and modelling process     

《地学前缘(英文版)》2020,11(6):2207-2219

This investigation assessed the efficacy of 10 widely used machine learning algorithms (MLA) comprising the least absolute shrinkage and selection operator (LASSO), generalized linear model (GLM), stepwise generalized linear model (SGLM), elastic net (ENET), partial least square (PLS), ridge regression, support vector machine (SVM), classification and regression trees (CART), bagged CART, and random forest (RF) for gully erosion susceptibility mapping (GESM) in Iran. The location of 462 previously existing gully erosion sites were mapped through widespread field investigations, of which 70% (323) and 30% (139) of observations were arbitrarily divided for algorithm calibration and validation. Twelve controlling factors for gully erosion, namely, soil texture, annual mean rainfall, digital elevation model (DEM), drainage density, slope, lithology, topographic wetness index (TWI), distance from rivers, aspect, distance from roads, plan curvature, and profile curvature were ranked in terms of their importance using each MLA. The MLA were compared using a training dataset for gully erosion and statistical measures such as RMSE (root mean square error), MAE (mean absolute error), and R-squared. Based on the comparisons among MLA, the RF algorithm exhibited the minimum RMSE and MAE and the maximum value of R-squared, and was therefore selected as the best model. The variable importance evaluation using the RF model revealed that distance from rivers had the highest significance in influencing the occurrence of gully erosion whereas plan curvature had the least importance. According to the GESM generated using RF, most of the study area is predicted to have a low (53.72%) or moderate (29.65%) susceptibility to gully erosion, whereas only a small area is identified to have a high (12.56%) or very high (4.07%) susceptibility. The outcome generated by RF model is validated using the ROC (Receiver Operating Characteristics) curve approach, which returned an area under the curve (AUC) of 0.985, proving the excellent forecasting ability of the model. The GESM prepared using the RF algorithm can aid decision-makers in targeting remedial actions for minimizing the damage caused by gully erosion.  相似文献   

海上风电升压站平台高程确定标准的合理性研究     

潘晓春 《海洋技术学报》2019,38(6):83-88

为保证海上风电升压电站建设的经济合理与安全可靠,合理确定海上风电升压电站平台高程十分必要。文中从波浪与潮位的遭遇组合、最大波高取值与现行相关标准的比较、最大波峰高度计算的合理性等方面,全面分析了确定海上风电升压站平台高程各组成项取值标准的合理性,研究认为现行标准明显偏高。建议海上升压站平台底部高程按"100年一遇极端高水位+重现期50年波列累积频率1%的最大波峰高度+安全超高"确定。结合工程实例计算分析,按本文建议可使海上升压站平台高程明显降低,从而节省工程造价,还可减轻升压站工程对周边风机的遮蔽影响,以达到多发电量的效果。  相似文献   

基于卷积神经网络与条件随机场方法提取乡镇非正规固体废弃物     

刘懿兰  黄晓霞  李红旮  柳泽  陈崇  王新歌 《地球信息科学学报》2019,21(2):259-268

随着村镇经济建设发展,生活垃圾和工业固体废弃物造成的污染问题日益突出,已经成为制约新农村建设发展和生态文明建设的关键问题,而目前针对乡镇非正规固体废弃物的调查与统计主要依赖全国各乡镇相关部门逐级调查上报,工作量较大。本文基于高分辨率遥感影像,将深度学习模型和条件随机场模型相结合引入到乡镇固体废弃物的提取研究中,探索一种基于深度卷积神经网络的乡镇固体废弃物提取模型。由于固体废弃物在影像上表现为面积小,分布破碎等特点,为了提高工作效率,将模型特分为识别和提取2个部分：① 通过全连接卷积网络（CNN）对固体废弃物进行快速识别判断,筛选感兴趣区域影像块;② 在传统的全卷积神经网络(FCN)的基础上加入条件随机场模型(CRF)提取固体废弃物边界,提高整体分割精度。根据安徽、山西等地区相关部门上报固体废弃物堆放点以及住房与城乡建设部城乡规划管理中心进行野外检查的结果,实验最终识别精度达到86.87%以上;形状提取精度为89.84%,Kappa系数为0.7851,识别与提取精度均优于传统分类方法。同时,该方法已经逐步应用于住房和城乡建设部有关成都、兰州、河北等部分乡镇非正规固体废弃物的核查工作,取得了较为满意的结果。  相似文献   

Effect of wave nonlinearity on fatigue damage and extreme responses of a semi-submersible floating wind turbine     

《Applied Ocean Research》2019

Floating wind turbine has been the highlight in offshore wind industry lately. There has been great effort on developing highly sophisticated numerical model to better understand its hydrodynamic behaviour. A engineering-practical method to study the nonlinear wave effects on floating wind turbine has been recently developed. Based on the method established, the focus of this paper is to quantify the wave nonlinearity effect due to nonlinear wave kinematics by comparing the structural responses of floating wind turbine when exposed to irregular linear Airy wave and fully nonlinear wave. Critical responses and fatigue damage are studied in operational conditions and short-term extreme values are predicted in extreme conditions respectively. In the operational condition, wind effects are dominating the mean value and standard deviation of most responses except floater heave motion. The fatigue damage at the tower base is dominated by wind effects. The fatigue damage for the mooring line is more influenced by wind effects for conditions with small wave and wave effects for conditions with large wave. The wave nonlinearity effect becomes significant for surge and mooring line tension for large waves while floater heave, pitch motion, tower base bending moment and pontoon axial force are less sensitive to the nonlinear wave effect. In the extreme condition, linear wave theory underestimates wave elevation, floater surge motion and mooring line tension compared with fully nonlinear wave theory while quite close results are predicted for other responses.  相似文献   

Deep neural network application for 4D seismic inversion to changes in pressure and saturation: Optimizing the use of synthetic training datasets     

Gustavo Côrte  Jesper Dramsch  Hamed Amini  Colin MacBeth 《Geophysical Prospecting》2020,68(7):2164-2185

In this work, we tackle the challenge of quantitative estimation of reservoir dynamic property variations during a period of production, directly from four-dimensional seismic data in the amplitude domain. We employ a deep neural network to invert four-dimensional seismic amplitude maps to the simultaneous changes in pressure, water and gas saturations. The method is applied to a real field data case, where, as is common in such applications, the data measured at the wells are insufficient for properly training deep neural networks, thus, the network is trained on synthetic data. Training on synthetic data offers much freedom in designing a training dataset, therefore, it is important to understand the impact of the data distribution on the inversion results. To define the best way to construct a synthetic training dataset, we perform a study on four different approaches to populating the training set making remarks on data sizes, network generality and the impact of physics-based constraints. Using the results of a reservoir simulation model to populate our training datasets, we demonstrate the benefits of restricting training samples to fluid flow consistent combinations in the dynamic reservoir property domain. With this the network learns the physical correlations present in the training set, incorporating this information into the inference process, which allows it to make inferences on properties to which the seismic data are most uncertain. Additionally, we demonstrate the importance of applying regularization techniques such as adding noise to the synthetic data for training and show a possibility of estimating uncertainties in the inversion results by training multiple networks.  相似文献   

Multi-band Solar Full-disk Extreme Ultraviolet Imager     

PENG Ji-long  FENG Tao-jun  SHI En-tao  LI Lin  YU Qian  ZHANG Kai  NIE Xiang-yu  MA Zi-liang 《Chinese Astronomy and Astrophysics》2019,43(3):444-456

Solar Extreme Ultraviolet (EUV) imaging observation is an important measure for the researches of solar activities and coronal plasma physics. But the traditional EUV imager and spectrograph can hardly achieve simultaneously the high spectral resolution and wide field-of-view of solar imaging. This paper has designed a new type of solar EUV multi-band imager, by adopting a kind of slitless grating and grazing incidence structure, it can realize the solar full-disk imaging of high spectral and spatial resolution. The field-of-view of the imager can be as broad as 47′. The spectral resolution is 2×10^?3nm per pixel, and the spatial resolution is 1.4′ per pixel. The temporal resolution of the solar full-disk is better than 60 s. The analysis of the solar full-disk spectral image and system response shows that the imager can observe the morphological evolutions of various solar activities, and can provide more comprehensive data for the researches of solar physics and space weather forecast.  相似文献   

基于支持向量机的京津冀城市群热环境时空形态模拟   总被引：1，自引：0，他引：1  

杨浩  孟娜  王婧  郑燕  赵莉 《地球信息科学学报》2019,21(2):190-200

城市群热环境作为区域生态重要组成部分,已成为近年来的研究热点。而如何选择针对城市群这种复杂地地貌特征的热环境量化工具一直是亟待解决的技术难点,基于此本研究提出了一种解决多样本、非线性、非平稳及高维函数拟合的计算方法,并建立了基于支持向量机(SVM)的京津冀城市群热环境曲面模型来揭示城市群热环境的时空形态变化。研究结果表明：① SVM模型在刻画多核心、多种土地利用类型城市群热环境的空间分布方面具有理论与实践可行性,能够根据热环境的整体空间布局通过高斯核函数进行局部优化差值,最大限度减少缺省值对模型拟合结果的影响。相比于对照方法可以模拟出更高精度的复杂地貌特征城市群热岛空间分布格局;② 在SVM模型曲面拟合的过程中,拟合精度和拟合时间是衡量拟合结果的重要指标,而原始影像的分辨率则是影响该指标的决定性因素;③ 2003-2013年区域内北京市与天津市的城市热岛效应变化最为明显,热岛面积分别增加7091 km²与4196 km²,空间上呈现出逐年接近连片发展趋势,热岛重心移动轨迹具有明显的时空分异性。北京城市热岛特征为东南部地区异速增长,西部地区缓慢增长;天津城市热岛特征为以城市中心为圆心向周围扩展。本研究进一步丰富了城市群热环境评测的定量方法,可以在实践上对城市群的城市规划、城市建设、环境保护和区域可持续发展等提供定量化、可视化的决策支持。  相似文献   

基于高分遥感影像的建筑物侧面信息提取及其高度计算     

宋旭东  曹文峰  冯德俊  张星星 《北京测绘》2020,(3):296-300

本文选取成都市某一区域建筑物A、B为研究对象,采用分辨率为0.61 m的Quick bird影像,运用图像分割法和LVQ神经网络算法,提取建筑物侧面信息,根据假设法原理,构建高度计算物理模型,求取建筑物高度。对比实测数据,结合可能影响实验结果的实地因素、遥感影像因素进行精度分析与评价,探讨基于高分遥感影像的建筑物侧面信息提取和高度计算的方法。结果表明,LVQ神经网络算法在建筑物侧面提取和高度计算中有更好的应用价值,精度高达94%。  相似文献   

A stochastic method for the prediction of icebreaker bow extreme stresses     

《Applied Ocean Research》2019

It is well established that the ship-ice interaction process is quite complex and associated ice loads on the icebreaker hull is a stochastic process. Obviously, novel accurate statistical methods and models should be developed and applied to estimate extreme bow stresses.This paper studies icebreaker bow stresses based on measured distribution of ice thickness in the Arctic Ocean on the way to and from the North Pole. Since the vessel route was carefully selected searching for easier ice conditions, the Arctic Ocean crossing was not a straight linear but a meandering path. Thus, the specific ship route data was biased with respect to general ice statistics in the region, but true with respect to the route specific ice data encountered by a ship navigating in that region. Therefore the route specific ice thickness data is directly needed for ship design and navigation analysis. It is assumed that captains are competent and knowledgeable, and therefore will select a route that provides the most favourable ice conditions.This paper contributes to study of the newest Chinese self-designed polar icebreaker, serving the purpose of enhancing icebreaker operational reliability. Finite Element Method software package ANSYS/LS-DYNA has been employed to simulate bow stress pattern for a particular icebreaker operating in the Arctic Ocean. Extreme bow stresses were estimated using Naess-Gaidai method. The latter is a first application of Naess-Gaidai method to a distribution with lower bound. Thus this paper aims at introducing an efficient method of estimating route-specific icebreaker extreme bow stresses.  相似文献   

Spatial landslide susceptibility assessment using machine learning techniques assisted by additional data created with generative adversarial networks     

Husam A.H.Al-Najjar  Biswajeet Pradhan 《地学前缘(英文版)》2021,12(2):625-637

In recent years,landslide susceptibility mapping has substantially improved with advances in machine learning.However,there are still challenges remain in landslide mapping due to the availability of limited inventory data.In this paper,a novel method that improves the performance of machine learning techniques is presented.The proposed method creates synthetic inventory data using Generative Adversarial Networks(GANs)for improving the prediction of landslides.In this research,landslide inventory data of 156 landslide locations were identified in Cameron Highlands,Malaysia,taken from previous projects the authors worked on.Elevation,slope,aspect,plan curvature,profile curvature,total curvature,lithology,land use and land cover(LULC),distance to the road,distance to the river,stream power index(SPI),sediment transport index(STI),terrain roughness index(TRI),topographic wetness index(TWI)and vegetation density are geo-environmental factors considered in this study based on suggestions from previous works on Cameron Highlands.To show the capability of GANs in improving landslide prediction models,this study tests the proposed GAN model with benchmark models namely Artificial Neural Network(ANN),Support Vector Machine(SVM),Decision Trees(DT),Random Forest(RF)and Bagging ensemble models with ANN and SVM models.These models were validated using the area under the receiver operating characteristic curve(AUROC).The DT,RF,SVM,ANN and Bagging ensemble could achieve the AUROC values of(0.90,0.94,0.86,0.69 and 0.82)for the training;and the AUROC of(0.76,0.81,0.85,0.72 and 0.75)for the test,subsequently.When using additional samples,the same models achieved the AUROC values of(0.92,0.94,0.88,0.75 and 0.84)for the training and(0.78,0.82,0.82,0.78 and 0.80)for the test,respectively.Using the additional samples improved the test accuracy of all the models except SVM.As a result,in data-scarce environments,this research showed that utilizing GANs to generate supplementary samples is promising because it can improve the predictive capability of common landslide prediction models.  相似文献