基于光谱—空间特征的ASTER影像岩性分类研究——以甘肃北山白峡尼山地区为例

遥感岩性制图是地质填图中的重要工作,基于光谱特征的岩性分类易受到色调、纹理等因素影响导致精度不佳。前人进行岩性自动分类研究多关注影像的光谱特征,而忽略空间特征,笔者等基于甘肃北山白峡尼山地区ASTER影像,将支持向量机、极限学习机两种机器学习分类方法与基于空间特征的快速漂移算法相结合进行岩性分类。结果表明支持向量机分类总体精度为89.17%;极限学习机不但具有需调节参数少的优势,且分类精度和速度均优于支持向量机,分类总体精度达96.70%;利用快速漂移算法提取的影像空间特征可有效减少错分区,提升岩性分类效果。研究证实将基于光谱特征的极限学习机和基于空间特征的快速漂移算法结合的岩性分类方法具有客观、高效、高精度等优势,可为后续地质填图和找矿勘查工作提供可靠数据支撑,在遥感岩性分类领域具有较高的推广价值。     

基于机器学习的锆石成因分类研究

锆石是在自然界中多种温压条件下能够稳定保存,并记录原岩年龄信息的副矿物。锆石微量元素能完整记录地质演化过程信息。通过微量元素分析锆石成因的研究已久,通常利用Th-U图解和La_N-(Sm/La)_N图解等二元图解对锆石进行分类研究。然而,随着锆石研究的深入,以及二元图解无法呈现数据高维度信息的局限性,传统图解已经不能满足对锆石类型进行准确判别,且对已知类型的锆石出现判定偏差。因此,本文将地质大数据与机器学习相结合,训练出高维度锆石成因分类器。文中收集了3 498条不同成因类型的锆石微量元素数据,并通过测试和运用随机森林、支持向量机、人工神经网络和k近邻等4种机器学习算法,最终得出准确率为86.8%的线性支持向量机锆石成因分类器,用于锆石类型的判定与预测。这项工作为锆石分类研究提供了更高维度的判别手段,极大提高了微量元素分析成因结果的精度。将锆石微量元素数据与机器学习方法相结合,是大数据分析与机器学习技术在地球化学研究中的积极探索。     

基于机器学习模型的淮河流域中长期径流预报研究

准确可靠的中长期径流预报是支撑水资源科学调配、提高水资源利用效率的关键。本研究采用AdaBoost模型(AdB)、随机森林模型(RF)和支持向量机模型(SVM)进行淮河流域王家坝和蚌埠站当年11月至次年10月共12个月的中长期径流预报研究。采用置换准确度重要性度量法从130项气象-气候因子及前期降雨/流量构建的1 562个因子变量中筛选出影响各月径流的关键因子,构建了基于AdB、RF和SVM模型的各月径流预报模型,模型参数采用随机搜索技术并结合交叉验证方式确定。采用变幅误差合格率和等级(五级)预报合格率指标对模型的预报精度进行了评估。变幅误差合格率指标表明,王家坝12个月的平均合格率分别为99.8%(AdB)、96.6%(RF)和95.9%(SVM),蚌埠站分别为100%(AdB)、94.8%(RF)和93.8%(SVM);等级预报合格率指标表明,王家坝12个月的平均合格率分别为79.0%(AdB)、76.4%(RF)和79.9%(SVM),蚌埠站分别为81.0%(AdB)、75.6%(RF)和76.6%(SVM)。模型均具有较好的预报效果,但RF和SVM模型对于高流量值的预报存在偏低现象,AdB模型整体上优于RF和SVM模型。     

基于机器学习的表层土壤成矿元素空间预测: 以稀有金属铷元素为例

近些年随着土地质量地球化学调查工作的开展,获取了大量表层土壤样品数据。然而,这些数据也存在一个明显的缺陷,即1∶50 000大比例尺表层土壤数据往往缺少成矿元素。鉴于土壤成矿元素含量对于矿产资源勘查的重要指示作用,尝试基于现有数据对大比例尺表层土壤成矿元素含量提供一个补全方案。以稀有金属铷元素为例,采用随机森林算法把同一区域2 548组1∶250 000小比例尺表层土壤数据按照8∶2的比例随机分为两组,用80%的数据进行训练建模,20%的数据对模型进行验证。采用变量重要性度量排序和构建学习曲线的组合方法优选了8种元素(K、B、Ni、V、Zn、As、Co、Cu)作为预测变量,模型对训练数据和测试数据的拟合优度R²分别达到0.983 2和0.895 6,说明预测变量的优选方法是有效的。随后将1∶50 000表层土壤的上述预测变量数据作为输入变量导入模型中,得到预测的Rb元素含量,预测结果比较符合实际特征。本研究表明将大数据机器学习随机森林算法引入表层土壤地球化学元素含量空间定量预测具有可行性,可进一步拓展土地质量地球化学数据的服务应用维度。     

A Bayesian network approach for predicting seismic liquefaction based on interpretive structural modeling

The Bayesian network (BN) is a type of graphical network based on probabilistic inference that has been gradually applied to assessment of seismic liquefaction potential. However, how to construct a robust BN remains underexplored in this field. This paper aims to present an efficient hybrid approach combining domain knowledge and data to construct a BN that facilitates the integration of multiple factors and the quantification of uncertainties within a network model to assess seismic liquefaction. Initially, only using given domain knowledge, a naive network model can be constructed using interpretive structural modeling. Thereafter, some effective information about the naive model is provided to construct a robust model using structural learning of BN from historic data. Finally, the returning predictive results and the predictive results are compared to other methods including non-probabilistic and probabilistic models for seismic liquefaction using the metrics of the overall accuracy, the area under the curve of receiver operating characteristic, prediction, recall and F₁ score. The methodology proposed in this paper achieved better performance, and we discussed the power and value of the proposed approach at the end of this paper, which suggest that BN is a good alternative tool for seismic liquefaction prediction.     

基于Faster R-CNN的野外环境中蝗虫快速识别

蝗虫是常见的害虫之一,对农作物和生态系统具有很大的危害,采用常规的方法对蝗虫进行监测存在一定局限性,为了有效应用海量野外影像数据实现对蝗虫实时监测,本文建立了一种基于深度学习网络的蝗虫自动识别模型。利用手机模拟摄像头获取的内蒙古锡林浩特附近草原的280张蝗虫的RGB图像,采用深度学习算法中的Faster R-CNN（Faster Region-based Convolutional Neural Network）网络结构建立了蝗虫识别模型。经验证该模型的精确度为0.756,可以较准确地将蝗虫从野外复杂环境中识别出来,与以往同类研究相比,在识别结果和实用性方面均有较大的进步。该模型是建立蝗虫实时监测系统的基础,可以为蝗虫的防治提供辅助信息,同时该网络结构还可以应用于其他害虫的识别,具有较强的推广性,拓宽了深度学习算法的应用领域。     

基于深度学习和雷达观测的华南短临预报精度评估

利用最新的深度学习算法,即卷积长短期记忆(Convolution Long-Short Term Memory)神经网络,构建基于深度学习的人工智能短临预报系统,以广州地区2019年3-5月雷达观测的数据为输入进行训练,然后进行短期1h内的降水预报。利用常用的统计评分指标(探测率POD、误报率FAR、临界成功指数CSI,相关系数CC)检验模型。结果表明,预报结果与实际观测的相关系数在1h内预报均保持在0.6以上,在1h内预报探测率均保持在80%以上,临界成功指数在降水强度为10mm·h^-1时,基本保持在60%,误报率均小于40%。     

21世纪高校GIS本科人才培养方案的研究

地理信息系统专业是一个新办专业，就目前而言，从教学计划、教学大纲、课程设置、教材的系统建设等方面都还存在问题。主要从GIS本科的培养目标计划、课程体系建设、实践教学及教与学的关系等方面阐述了思想，以与大家探讨如何建设一个规范、成熟的地理信息系统专业。     

Self-perceived Roles in Life and Achieving Sustainability on Family Farms in North-eastern Australia

Rural production landscapes in Australia are experiencing a rapid rate of change as a result of, among other factors, climate change, biodiversity loss and changing societal values. Consequently, there is increasing pressure on producers to increase their sustainability. Understanding how producers perceive themselves in the context of this changing landscape is limited but important for the design of policy effective for achieving sustainability. This paper is based on a case study in the north-eastern Australian rangelands that included face-to-face interviews with 28 beef producers and a telephone survey with another 91 producers. The study investigated male and female beef producers' self-perceived roles in life through a lens of different farming discourses and the relationship between these roles and beliefs aligned with sustainability. Results revealed that although producers' self-percieved roles in life were being constructed through a mix of more or less ‘traditional’ discourses, tradition was still a strong influence. Producers who strongly identified with roles linked to ‘less traditional’ discourses were more likely than those who strongly identified with production-orientated roles to agree with beliefs that favoured nature conservation, learning and adapting to change. Increased opportunities for producers to participate in alternative discourses would appear important for fostering a self-identity that is open to learning, difference and change.     

The Utility and Organization of a College Field Course: Examining National Park Management

Abstract

This article reviews the benefits of field-based instruction, particularly as it complements traditional classroom learning. It suggests that direct observation and hands-on learning associated with field-based instruction raise students' interest level and improve their understanding and long-term retention of targeted concepts. In addition, planning, implementation, and evaluation strategies are provided and recommendations are outlined for instructors interested in organizing or improving a field course at their learning institution. The park management field course offered between spring and summer semesters at Oklahoma State University provides an example of the major components of a successful field course.