一种基于GPU和内存映射文件的高分辨率遥感图像快速处理方法

高分辨率遥感图像处理经常面临程序执行时间过长和内存空间不足的问题,虽然并行计算技术可以提高遥感图像的处理速度,但是无法降低算法占用的巨大内存空间。为了解决这一问题,本文提出了一种利用CUDA和内存映射文件的高分辨率遥感图像快速处理方法,并以K-Means算法为例进行了实现。其中,CUDA技术可以有效利用GPU强大的并行计算能力,而内存映射文件技术降低了磁盘I/O速度较慢对算法性能的影响。实验结果表明,本文方法比传统K-Means聚类算法计算速度提高了30倍左右,内存使用量降低了90%以上。     

Rare earth elements and yttrium in ferromanganese deposits from the South China Sea: distribution, composition and resource considerations

Ferromanganese nodules and crusts contain relatively high concentration of rare earth elements(REE) and yttrium(REY),with a growing interest in exploitation as an alternative to land-based REY resources.On the basis of comprehensive geochemical approach,the abundance and distribution of REY in the ferromanganese nodules from the South China Sea are analyzed.The results indicate that the REY contents in ferromanganese deposits show a clear geographic regularity.Total REY contents range from 69.1×10~(–6) to 2 919.4×10~(–6),with an average value of 1 459.5×10~(–6).Especially,the enrichment rate of Ce content is high,accounting for almost 60% of the total REY.This REE enrichment is controlled mainly by the sorption of ferromanganese oxides and clay minerals in the nodules and crusts.Moreover,the total REY are higher in ferromanganese deposits of hydrogenous origin than of diagenetic origin.Finally,Light REE(LREE) and heavy REE(HREE) oxides of the ferromanganese deposits in the study area can be classified into four grades: non-enriched type,weakly enriched type,enriched type,and extremely enriched type.According to the classification criteria of rare earth resources,the Xisha and Zhongsha platform-central deep basin areas show a great potential for these rare earth metals.     

南海北部冷泉区深水珊瑚的发现及其特征

冷水珊瑚也称为深水珊瑚, 在生物多样性、生态资源和科研价值等方面具有重要意义。文章对采自南海北部冷泉区的冷水珊瑚骨骼碎屑进行测定, 鉴定出冷水珊瑚2个种(Crispatotrochus sp.1和Crispatotrochus sp.2), 以及4个属[Balanophyllia (Balanophyllia)、Balanophyllia (Eupsammia)、Lochmaeotrochus和Enallopsammia]。测定的冷水珊瑚的δ¹³C为-7.36‰~-1.15‰, δ¹⁸O为-1.38‰~3.67‰, 与全球冷水珊瑚碳氧同位素组成相似, 但明显不同于南海暖水珊瑚、冷泉碳酸盐岩及低温热液成因碳酸盐岩的碳氧同位素组成。     

南海北部陆坡深水区浅层天然气藏特征

南海北部陆坡深水区的浅层天然气藏是一种伴随天然气水合物的新型油气藏, 具有埋藏浅、规模大的特点, 其埋藏深度一般小于300m。浅层天然气藏由深部裂解气沿断裂上升被天然气水合物封盖而形成, 识别似海底反射(BSR)是寻找浅层天然气藏有效方法。浅层天然气藏的气源主要有热解气、生物气和混合气, 陆坡张性断裂是气体运移的主要通道, 水合物下部的砂层是浅层天然气藏的主要储集层, 水合物层则是封盖层。从南海发现的天然气水合物分布特征看, 浅层天然气藏在陆坡深水区广泛分布且气藏厚度大, 潜在资源量非常可观, 是一种新型的开采成本相对低廉的油气藏。     

An algorithm to resolve γ-rays from charged cosmic rays with DAMPE

The DArk Matter Particle Explorer(DAMPE),also known as Wukong in China,which was launched on 2015 December 17,is a new high energy cosmic ray and γ-ray satellite-borne observatory.One of the main scientific goals of DAMPE is to observe Ge V-Te V high energy γ-rays with accurate energy,angular and time resolution,to indirectly search for dark matter particles and for the study of high energy astrophysics. Due to the comparatively higher fluxes of charged cosmic rays with respect to γ-rays,it is challenging to identify γ-rays with sufficiently high efficiency,minimizing the amount of charged cosmic ray contamination. In this work we present a method to identify γ-rays in DAMPE data based on Monte Carlo simulations,using the powerful electromagnetic/hadronic shower discrimination provided by the calorimeter and the veto detection of charged particles provided by the plastic scintillation detector. Monte Carlo simulations show that after this selection the number of electrons and protons that contaminate the selected γ-ray events at～10 Ge V amounts to less than 1% of the selected sample.Finally,we use flight data to verify the effectiveness of the method by highlighting known γ-ray sources in the sky and by reconstructing preliminary light curves of the Geminga pulsar.     

Preprocessing photospheric vector magnetograms for nonlinear force-free field extrapolation of the global corona

Nonlinear force-free magnetic field(NLFFF) extrapolation based on the observed photospheric magnetic field is the most important method to obtain the coronal magnetic field nowadays.However, raw photospheric magnetograms contain magnetic forces and small-scale noises, and fail to be consistent with the force-free assumption of NLFFF models. The procedure for removing the forces and noises in observed data is called preprocessing. In this paper, we extend the preprocessing code of Jiang Feng to spherical coordinates for a full sphere. We first smooth the observed data with Gaussian smoothing, and then split the smoothed magnetic field into a potential field and a non-potential field.The potential part is computed by a numerical potential field model, and the non-potential part is preprocessed using an optimization method to minimize the magnetic forces and magnetic torques. Applying the code to synoptic charts of the vector magnetic field from SDO/HMI, we find it can effectively reduce the noises and forces, and improve the quality of data for a better input which will be used for NLFFF extrapolations applied to the global corona.     

Effects of feeding time on complement component C7 expression in Pelteobagrus vachellii subject to bacterial challenge

Journal of Oceanology and Limnology - Shifting the feeding time for fish from daytime to nighttime could alter their digestive behavior, disturb their metabolism, and may affect immune-related...     

基于车载激光点云的街景立面自动提取

街道景观图是城市规划设计和城市管理的重要参考依据,车载点云数据能够提供沿街建筑的三维点信息,精度高,覆盖范围广泛,为街景立面整治提供了新的解决方案。为此,本文提出一种适用于车载点云的街景立面的自动提取方法,提取立面点云的具体步骤为：对原始数据去噪滤波;选取非地面点构建规则格网并二值化,依据语义特征筛选出建筑物点云;用POS数据拟合直线段帮助选取参考向量与参考平面;计算点云到参考面的距离,按距离分类点云数据,并对前述步骤中未分类点另行提取,合并面点集得到以沿街建筑物立面为主的街景立面点云。为了验证这一方法的可行性和有效性,采用点云数据进行实验,实验结果表明本方法在一定程度上提高了数据处理效率,能得到较理想的结果。     

1.5 μm波长n-p-n型InGaAsP-InP晶体管激光器材料外延生长

基于器件模拟仿真,设计了一种1.5 μm波长InGaAsP-InP晶体管激光器材料外延结构. 其多量子阱有源区置于基区非对称波导中. 仿真结果显示该外延结构能够获得较好的光场限制和侧向电流限制. 对该材料MOCVD生长研究表明,基极重掺杂接触层中Zn²⁺扩散将导致量子阱严重退化. 通过对其扩散过程的模拟仿真,采用平均掺杂浓度为1×10¹⁸ cm^-3的梯度掺杂,有效地抑制了Zn²⁺向量子阱区的扩散. 所获得的外延材料在     

ATON型Hall推力器缓冲区预电离问题研究

ATON是20世纪90年代新提出的一种Hall推力器设计方案,其中的缓冲区是ATON新采用的一种结构.本文对缓冲区的作用进行分析,在此基础上提出了采用附加电源提高缓冲区预电离率的方法.并采用PIC(Particle-in-cell)粒子模拟方法对这种设计方式的等离子体分布进行数值模拟.结果表明,通过增加缓冲区内的附加电压,能够有效地提高缓冲区预电离率.