江苏沿海滩涂利用模式遥感影像解译和时空变化分析

为了弄清江苏沿海滩涂利用模式分布情况,定量分析其时空变化趋势,本文基于Google Earth Engine云计算平台和Landsat卫星遥感影像数据,结合随机森林分类方法,通过机器学习建立了滩涂利用模式自动识别分类方法,样本训练后识别准确率达93％以上.利用该方法提取了2002年、2014年和2019年江苏沿海的土地...     

合浦珠母贝贝壳珍珠层EDTA可溶性基质蛋白对无定形碳酸钙稳定性的影响

通过红外光谱、扫描电镜、能谱等方法探讨了合浦珠母贝贝壳珍珠层EDTA可溶性基质蛋白(ESM)对无定形碳酸钙(ACC)晶型转化的影响。试验结果显示,合浦珠母贝贝壳珍珠层ESM在镁离子存在的条件下能够显著抑制ACC的转化,并减少最终诱导形成的方解石中的镁离子含量。同时,珍珠层ESM具有调节方解石晶体晶貌的作用。     

Precise and Accurate In Situ Determination of Lead Isotope Ratios in NIST,USGS, MPI‐DING and CGSG Glass Reference Materials using Femtosecond Laser Ablation MC‐ICP‐MS

Lead isotope ratio data were obtained with good precision and accuracy using a 266 nm femtosecond laser ablation (fLA) system connected to a multi‐collector ICP‐MS (MC‐ICP‐MS) and through careful control of analytical procedures. The mass fractionation coefficient induced by 266 nm femtosecond laser ablation was approximately 28% lower than that by 193 nm excimer laser ablation (eLA) with helium carrier gas. The exponential law correction method for Tl normalisation with optimum adjusted Tl ratio was utilised to obtain Pb isotopic data with good precision and accuracy. The Pb isotopic ratios of the glass reference materials NIST SRM 610, 612, 614; USGS BHVO‐2G, BCR‐2G, GSD‐1G, BIR‐1G; and MPI‐DING GOR132‐G, KL2‐G, T1‐G, StHs60/80‐G, ATHO‐G and ML3B‐G were determined using fLA‐MC‐ICP‐MS. The measured Pb isotopic ratios were in good agreement with the reference or published values within 2s measurement uncertainties. We also present the first high‐precision Pb isotopic data for GSE‐1G, GSC‐1G, GSA‐1G and CGSG‐1, CGSG‐2, CGSG‐4 and CGSG‐5 glass reference materials obtained using the femtosecond laser ablation MC‐ICP‐MS analysis technique.     

Matrix‐rich and associated matrix‐poor sandstones: Avulsion splays in slope and basin‐floor strata

A common facies observed in deep‐water slope and especially basin‐floor rocks of the Neoproterozoic Windermere Supergroup (British Columbia, Canada) is structureless, coarse‐tail graded, medium‐grained to coarse‐grained sandstone with from 30% to >50% mud matrix content (i.e. matrix‐rich). Bed contacts are commonly sharp, flat and loaded. Matrix‐rich sandstone beds typically form laterally continuous units that are up to several metres thick and several tens to hundreds of metres wide, and commonly adjacent to units of comparatively matrix‐poor, scour‐based sandstone beds with large tabular mudstone and sandstone clasts. Matrix‐rich units are common in proximal basin‐floor (Upper Kaza Group) deposits, but occur also in more distal basin‐floor (Middle Kaza Group) and slope (Isaac Formation) deposits. Regardless of stratigraphic setting, matrix‐rich units typically are directly and abruptly overlain by architectural elements comprising matrix‐poor coarse sandstone (i.e. channels and splays). Despite a number of similarities with previously described matrix‐rich beds in the literature, for example slurry beds, linked debrites and co‐genetic turbidites, a number of important differences exist, including the stratal make‐up of individual beds (for example, the lack of a clean sandstone turbidite base) and their stratigraphic occurrence (present throughout base of slope and basin‐floor strata, but most common in proximal lobe deposits) and accordingly suggest a different mode of emplacement. The matrix‐rich, poorly sorted nature of the beds and the abundance and size of tabular clasts in laterally equivalent sandstones imply intense upstream scouring, most probably related to significant erosion by an energetic plane‐wall jet or within a submerged hydraulic jump. Rapid energy loss coupled with rapid charging of the flow with fine‐grained sediment probably changed the rheology of the flow and promoted deposition along the margins of the jet. Moreover, these distinctive matrix‐rich strata are interpreted to represent the energetic initiation of the local sedimentary system, most probably caused by a local upflow avulsion.     

基于最大特征值估计的C3算法及应用

第三代相干体算法(C3算法)具有分辨率高、压制噪声能力强的优点,但是该算法需要计算协方差矩阵的特征值,所以耗时较多。为提高大矩阵特征值分解的运算速度,提出了一种估计最大特征值的快速收敛算法,通过运用A2k矩阵的迹来构造一种收敛速度更快的特征值估计方法,并给出了这种方法的误差估计表达式,实现计算精度的有效控制。文中还讨论了地震数据体边缘补地震道方法,解决了将递推算法应用到平面方向时空间窗越过边界的问题。实例表明,本算法的计算效率明显高于常规的C3算法。     

嫦娥3号巡视探测器桅杆的转动角度误差检校

嫦娥3号巡视探测器桅杆的展开机构、偏航机构和俯仰机构在转动过程中存在系统性的偏差,这些误差影响了巡视探测器导航定位、地形恢复和路径规划的精度.针对这一问题,文章提出了一种间接平差模型,通过真实数据试验证明该方法能够用于对嫦娥3号巡视探测器桅杆的转动角度误差进行检校,并且具有很高的精度和稳定性.     

基于布尔矩阵的空间关联规则提取方法研究

空间关联规则是空间数据挖掘(SDM)中的重要内容之一。由于空间数据的复杂性,传统的空间关联规则挖掘方法主要是将空间数据库变换为非空间数据库,通过挖掘算法挖掘空间关联规则。目前,Apriori算法是关联规则挖掘中使用最为普遍的算法,但是,由于该算法在关联规则提取过程中需要多次扫描数据库,并且产生冗余的候选项集,因此,在执行大型数据库的关联规则挖掘时,具有效率低下的缺陷。本文基于Apriori算法提出了基于布尔矩阵的空间关联规则挖掘算法,并以挖掘福建省厦门市土地覆盖现状与地形特征因子的空间关联关系作为试验案例,对比Apriori算法的提取结果与提取效率,结果表明:该算法不仅减少了扫描数据库的次数,而且减少了冗余候选项集的产生,提高了空间关联规则的提取效率。     

Intel MKL在Bernese GNSS数据处理中的应用

为提高Bernese GNSS software数据处理效率，将英特尔数学核心函数库(math kernel library，MKL)应用于Bernese 精密定轨数据处理，对比分析多个MKL矩阵求逆函数与Bernese 原有程序的计算效率。使用2019-03全球200个测站北斗/GNSS数据进行实验分析，结果表明，采用参数预消除策略时，参数预消除步骤消耗时间明显大于矩阵求逆，使用MKL处理数据效率提升不明显；而未采用参数预消除策略时，使用MKL矩阵求逆函数可显著提高矩阵求逆效率，其中dpotri函数矩阵求逆计算效率最高，消耗时间平均值为133 s，相比Bernese 原有程序计算速度可提高13倍。     

用传输矩阵方法研究多层介质增透膜的透射特性

利用麦克斯韦方程组和边界条件,推导出由高低折射率介质组成周期性的多层增透膜系的传输矩阵,用传输性矩阵求出增透膜系的反射率和透射率表示式,采用Maple语言和数值计算方法得到4层和8层高反射膜系的反射率随入射光波长变化曲线,分析和讨论入射光的入射角对反射率和透射率的影响.得到的结果有利于对多层介质增透膜系结构的理解,为光学仪器镀膜的设计提供了理论依据.     

基于弱势人群个体可达性评价的社区服务设施配置研究

当前我国人口形势面临老龄化与少子化风险,抓好“一老一小”、确保老有所养和幼有所育成为我国城乡发展中特别关注的方面。在这一背景下,如何营造有利于养老和生育的社会环境,如何针对老年人、育龄妇女和儿童等弱势人群提供更为合理的设施配置成为民生工程的重要内容。社区服务设施作为城市公共服务的空间载体,其配置模式的完善与提升是落实以人为本、集中体现社会公平的重要路径。当前,从网络地图规划路径API获取的出行时间成本矩阵能够为可达性研究提供接近真实的出行时间数据,此外个体化的人口数据如实有人口数据能够以其丰富的属性信息为识别各类弱势人群提供依据,同时精确刻画人口的空间分布。本文基于弱势人群的需求特征与相关规范,对老年人、育龄妇女和儿童三类弱势人群提出了明确定义,并构建了面向各类弱势人群个体使用需求的社区服务设施供给标准。在此基础上,利用实有人口数据、网络地图出行时间成本矩阵和POI数据,综合考虑社区服务设施的服务容量,提出了基于弱势人群个体可达性评价的社区服务设施供给评价与布局优化方法,并以广州市人民南社区为例进行了应用实践。研究表明,人民南社区的弱势人群服务设施供给存在不足,优化配置后设施的供需情况更加合理,服务人口覆盖率亦明显提升。本文提出的方法能够从独立个体尺度精确评价社区弱势人群服务设施的布局合理性,在拓展数据类型及应用方法方面为当前人口形势下精细化研究服务设施的空间分布和供需问题提供新的思路。