Episodic fluid flow as a trigger for Miocene‐Pliocene slope instability on the Utgard High,Norwegian Sea

Mass wasting is triggered on many continental slopes by a number of mechanisms, including seismic shaking, high sedimentation rates, the presence of weak geological units and gas hydrate dissociation. In this study, the morphology of a Late Miocene–Early Pliocene mass‐transport complex (MTC) on the Utgard High is unravelled and discussed in relation to possible trigger mechanisms. The approach used here includes 3D seismic interpretation and the analysis of variance attribute maps. The interpreted MTC is located on the crest and flanks of the Utgard High and is composed of three mass‐transport deposits with seismic characters varying from transparent and chaotic seismic facies at the base to slightly deformed layers composed of mounds and rafted blocks in the middle and chaotic to transparent reflections at the top. Lithologically, the MTC consists predominantly of claystone with high gamma ray and low density and resistivity values, demonstrating that the associated mounds represent remobilized ooze sediments. A vertical stack of six magmatic sills emplaced from 55.6 to 56.3 Ma into the Upper Cretaceous shales is interpreted at depths of 3,000–5,500 ms two‐way travel time (TWTT). In association with these magmatic sills are several hydrothermal vent complexes that interacted with the top MTC horizon, signifying that episodic and secondary fluid‐venting events might be the principal mechanism facilitating mass wasting in the study area. In addition, the remobilization of ooze sediments into mounds is hypothesized to be dependent on fluids and clayey layers. As a corollary of this work, the importance of relict and recurrent episodes of fluid flow in the Vøring Basin and their influence on the geotechnical integrity of the overburden and later mass wasting is established.     

Research on the Development Status and Strategy of Productive Rooftop Gardens on Commercial Complexes

The introduction of productive gardens on public building roofs is an active way to use urban idle space. It has ecological, economic, and social values and helps alleviate many urban problems caused by the rapid advancement of land urbanization. This paper takes the productive rooftop garden of an urban commercial complex as an example, and assesses its development status based on methods including ArcGIS, field research, and questionnaire interviews, combined with the overall aspects of the country and key case analysis. The results indicate several key aspects of the current status of such gardens in China. (1) As affected by natural and social factors, the current spatial distribution of productive rooftop gardens of commercial complexes in Chinese mainland is uneven, with 84.21% located in the southeast coast and the Sichuan region. (2) The operation and development of this type of productive landscape is in good shape. The number of rooftop gardens has continued to increase since 2013, and the scale is generally greater. Currently, the business model which combines nature education and parent-child amusement experience activities is the most stable. (3) Cases in good operating condition tend to have relatively related characteristics in layouts, traffic functions, landscape elements, and space design. (4) Questionnaire interviews show that citizens are highly willing to participate in rooftop productive landscapes, while operators still experience challenges in policies, funds, and planting knowledge in practice. This paper analyses the existing problems in the development status and strategy of the rooftop productive landscape. It proposes complementary optimization strategies to serve as a reference for the rooftop design of commercial complexes and the utilization of a significant amount of idle space on the roofs.     

Navigating the Complex Trade-Offs of Pesticide Use on Santa Cruz Island,Galapagos

Pesticide use by farmers on Santa Cruz Island, Galapagos—a perceived adaptation to changing ecological and economic dynamics—has the potential to lead to environmental degradation in an area that is known and valued worldwide for its biodiversity. We survey Santa Cruz farmers to understand motivations for and concerns about pesticide use on the island. Results from farmer surveys are supplemented with interview data to develop the case study of pesticide use on Santa Cruz Island. We then apply a “complex trade-off” framework to explore and navigate the tensions between conservation and livelihoods. We conclude by elaborating the implementation of a participatory certification system, the Participatory Guarantee System, as a possible path for reconciling trade-offs in Santa Cruz, Galapagos.     

结合模平方的双树复小波变形监测数据滤波

针对变形监测数据的去噪问题,该文在分析离散小波变换去噪不足的基础上,提出了一种基于模平方的双树复小波变形监测数据滤波方法。该方法利用双树复小波变换的完全重构、近似平移不变性和较好的方向选择性等特点,通过最小尺度空间的小波系数得到噪声强度,并结合模平方处理法确定各层的阈值,经重构阈值处理后的各层小波系数即得到去噪后的信号;经算例,并与传统离散小波变换对比分析。结果表明:双树复小波变换的分解效果优于传统离散小波变换,能较好地表现出细节部分的频率信息,使变形信号的周期性变化特征更为明显。该方法去噪更彻底,进一步提高了消噪的精度和可靠性,可作为变形监测数据降噪处理的新方法。     

极化干涉SAR森林树高提取算法的对比分析

在对极化干涉SAR森林树高反演的DEM差值算法、相干相位-幅度综合反演算法进行分析的基础上,对基于极化干涉相干优化方法的改进算法进行了探讨。利用黑龙江大兴安岭地区的一对ALOS全极化干涉SAR数据进行实验,并对比分析各算法的反演结果。结果表明,在使用改进的算法进行森林树高反演时可以获取精度较高的反演结果,并且在一定程度上提高了森林树高反演的稳定性,为森林树高反演工作的业务化运行提供一定的依据。     

复杂城区车载激光扫描点云分层路段定位与数据纠正实验研究

高精度的车载点云数据是实现各种城市地物要素提取的前提,但在复杂城市环境下车载作业时信号遮挡、衰减和多径效应频繁发生,进而造成点云数据精度严重降低,如何对复杂环境下的点云数据进行分析和纠正就显得至关重要.现有的质量分析方法多从测距误差、仪器安装误差和数据处理误差方面出发,虽然可以在一定程度上改化数据,但在如何快速定位出分...     

干涉合成孔径雷达（INSAR）复图像配准方法

合成孔径雷达干涉测量技术是使用图像中的相位信息来提取高精度的数字地面高程模型。精确的图像配准是合成孔径雷达干涉技术提取数字高程模型的关键步骤之一，配准质量将直接影响提取的地面高程的精度。配准按照其精度分为粗配准和精配准。文中采用基于幅度互相关函数法完成图像的粗配准。在粗配准的基础上，使用最小二乘配准算法进一步提高配准的精度。实际数据处理表明，配准精度能够满足INSAR数据处理的要求。     

基于复小波变换的SAR图像斑点滤波方法

斑点噪声的存在使得SAR图像的应用受到较大的限制.文中首先介绍了SAR图像斑点噪声模型以及几种常用的SAR斑点噪声抑制方法;然后对小波变换斑点噪声滤波方法进行分析,提出了基于二元树复小波变换的SAR图像滤波方法.实验比较表明,二元树复小波变换抑制SAR斑点噪声效果明显,较小波变换更具有优越性.     

复杂卫星图像中的小目标船舶识别

船舶作为海上的重要目标,实现对船舶自动识别有重要的意义。针对卫星图像中云雾、海岸背景等复杂海情对船舶识别带来的干扰,以及小目标船舶高漏检率问题,本文提出一种多尺度深度学习模型训练策略,在此基础上构建了一种船舶识别的深度学习网络,该网络可分为多尺度训练、特征提取、生成目标建议区域、船舶分类这4个部分。首先,采用多尺度的训练策略,将多尺度的船舶样本送入网络中进行训练,这样在训练样本中加入了大量小目标船舶的样本,使网络充分提取到小目标船舶的特征;其次,通过卷积神经网络对目标船舶进行特征自适应提取;然后,目标区域建议网络可依据卷积神经网络提取到的特征,在图像中找到感兴趣目标区域,即框定船舶的位置;最后,通过多个全连接层的组合,将高维特征映射到一个4元组中,再运用分类函数输出每一类船舶的概率值,概率值最大的则为该船舶的类别。同时为解决云雾遮挡和海岸背景的干扰,采用了一种负样本增强学习的方法,在样本数据集中加入了大量只含有云雾和海岸背景的图片,进行负样本扩充,增强网络模型对云雾及海岸背景的特征学习能力,以此解决复杂海情的影响。实验结果表明,所提方法有效解决了复杂海情条件下的船舶识别难,以及小目标船舶识别难的问题,实现了复杂海情条件下的船舶识别。同时,与现有成熟的深度学习目标识别算法相比,本文算法的精确度和召回率分别提升了6.98%和18.17%,所训练的模型具有良好的泛化能力和鲁棒性。     

规则池塘群的复杂直线模式识别

针对现有面要素直线模式识别方法难以解决规则池塘中破碎区域的问题,提出了一种规则池塘群的复杂直线模式识别方法。首先,分析了规则池塘群的空间特征和认知特点,提出了“主次关系→并列关系→直线模式→复杂直线模式”的多层次认知顺序;其次,设计了主次关系池塘组、并列关系池塘组、直线模式池塘群的识别方法;最后,构建了复杂直线模式池塘群的识别模型。试验表明:本文方法能够消除破碎区域对规则池塘群直线模式提取的不利影响,有效提升复杂直线模式识别的质量。