Atmospheric trace gas analysis using matrix isolation-fourier transform infrared spectroscopy

A novel cryogenic sampling method combining the matrix isolation technique with FTIR spectroscopy has been developed for atmospheric trace gas analysis. It is applicable to a wide range of molecules with detection limits typically in the 10–50 ppt range. The method is described along with some measurements of N₂O, CFCl₃, CF₂Cl₂, OCS, CS₂, SO₂ and PAN from samples collected at ground level and from an aircraft between 9 and 14 km.     

石墨炉原子吸收光谱法测定血液中微量铅

研究了石墨炉测定血液中微量铅的适宜条件,用氘灯校正背景,磷酸二氢铵作基体改进剂,方法具有准确、快速、简便等优点,用于扬州妇幼保健院实际样品的测定,结果预示我们的环境不容忽视,血液中铅含量能够比较真实地反映生物个体受到的污染程度。     

Foundation pit dewatering optimization design based on GMW-2005 and LGR technique

The slope instability and uneven settlement caused by the foundation pit dewatering pose a great threat to the project and surrounding environment.Thus,it is necessary to carry out optimization design of the foundation pit dewatering project.To solve this problem,this paper introduces the simulation-optimization method and establishes the groundwater model based on MODFLOW-2005.The local grid refinement(LGR)technique is employed to achieve the refinement of dewatering area.Under the premise of construction safety,the minimum cost of the project is set as the objective function to solve the optimization problem of foundation pit engineering with GWM-2005.The results show that the optimization with GWM-2005 will be more accurate combined with the LGR model,but the improvement of the optimization is not obvious.It is necessary to choose a suitable local refinement model considering the engineering requirement.     

垂直井眼中三维阵列感应测井资料自适应井眼校正

三维感应测井受井眼环境影响很大,如何消除这些不利影响是有效获取各向异性地层电导率真值并提高资料处理和解释质量的一项重要工作.本文以一维柱状各向异性地层模型为基础,通过最小平方反演技术从实际测量资料中快速反演井径、偏心距、泥浆电导率、地层水平电导率和各向异性系数等模型参数,设法减小井眼环境对测量结果的影响.首先,利用仪器偏心情况下不同旋转角电导率张量的关系方程,建立旋转角提取与校正方法,获取无旋转角三维感应测井数据.在此基础上,根据事先计算好的井眼校正库并结合多维非线性有限元逼近技术,建立快速计算三维感应测井响应以及Jacobi导数矩阵的插值公式,并利用自适应正则化迭代反演技术不断修改模型参数,实现理论合成资料与输入资料的最佳拟合.最后,通过反演出的模型参数计算三维感应测井资料所有分量的校正量,实现三维感应资料的井眼校正.理论模型和实际资料的处理结果均证实,在低阻井眼泥浆情况下,一维柱状模型中的反演方法仍然可以提取出较可靠的地层水平和垂直电导率,且井眼环境校正对于受井眼环境影响相对较小的测井曲线具有非常好的校正效果.     

An extension of general identities for 3D water-wave diffraction with application to the Diffraction Transfer Matrix

Interaction theories are used in numerous branches of physics to efficiently evaluate wave scattering by multiple obstacles. An example of these interaction theories is the direct matrix method introduced by Kagemoto and Yue [1], which enables fast computation of three-dimensional water-wave multiple-scattering problems. The building block of interaction theories is a mathematical operator that encapsulates the mapping between incident and scattered waves. This operator is generally referred to as T-matrix and satisfies both reciprocity and energy identities. In some branches of physics, such as acoustics and electromagnetism, these identities are well established; in hydrodynamics, however, they have only been derived for a T-matrix that maps two-dimensional incident and scattered water waves. In three dimensions, water waves can be represented as a series expansion of cylindrical eigenfunctions. In this paper, we use this representation of water waves to derive the reciprocity and energy identities satisfied by the T-matrix of the direct matrix method, known as Diffraction Transfer Matrix (dtm). The identities derived herein represent an extension of existing general relations between two diffraction solutions. We show that this extension can be applied to verify the accuracy of the dtm entries, thereby increasing the reliability of existing schemes for computing the dtm. We present results for the dtm of two geometrically different isolated obstacles, as well as for the dtm of an asymmetric array. Finally, we demonstrate that the results presented herein can be extended to floating bodies found in a wide range of ocean engineering problems.     

2021年南京市新冠疫情期间大气污染物变化特征及来源解析

为了解2021年南京市新冠疫情期间城市大气污染物浓度的变化和成因,利用南京大学SORPES站点2021年7月1日—2021年8月30日大气污染物在线监测数据,分析疫情前、中、后颗粒物及气态污染物的浓度变化,针对臭氧(O₃)的关键前体物,挥发性有机物(Volatile Organic Compounds,VOCs)采用正定矩阵因子分解法模型(Positive Matrix Factorization,PMF)、拉格朗日粒子输送与扩散模型(Lagrangian Particle Distribution M odel,LPDM)分析其污染来源。结果表明:疫情封闭期间,南京市PM_2.5质量浓度较疫情前降低了40%～50%,组分中硝酸盐、有机物质量浓度降幅最为显著,分别下降了34.0%和16.5%。臭氧体积浓度不降反升,城中区域增幅最显著站点可达50%左右。其气态前体物氮氧化物(NO_x)及VOCs浓度变化呈相反态势,分别较疫情前降低28%、升高49.6%。模型及卫星遥感结果表明,疫情期间南京市臭氧属于协同偏VOCs控制区。气团溯...     

利用多频观测数据快速解算整周模糊度的算法研究

本文首先阐述了多频观测数据解算模糊度时所用到的矩阵转换的一些方法,讨论了由矩阵变换方法和CAR方法可以引中出的模糊度解算方法,指出TCAR/MCAR方法模糊度解算精度低的原因并改进,对短基线利用TCAR/MCAR方法进行模糊度解算,并与矩阵变换法做了比较.     

Ecological performance and possible origin of a ubiquitous but under-studied gastropod

Invasions by non-indigenous species (NIS) have been suggested to alter local, regional and global biota on unprecedented scales. To manage NIS, it is pivotal to identify whether a species is introduced or native, but even today the geographical origin of thousands of species worldwide remain uncertain. Most of these ‘cryptogenic species’ are inconspicuous and rare, but in a few instances, they can also be abundant and conspicuous species, with large impacts on community structure. The identification of cryptogenic species, and summarizing information on their most likely origin, is an important task in invasion biology, and can highlight the need for research and management. Here, we document that the gastropod Batillaria australis in the Swan River estuary (Perth, Western Australia) is a conspicuous species of uncertain origin. A literature review combined with new survey data revealed that all evidence point to a recent human-mediated transfer; for example, it is absent from the fossil record, was first collected in 1954, has a low parasite diversity, has increased its population size dramatically in recent times, is separated by >3000 km from conspecifics, has no long-distance dispersal mechanisms, and existing ocean currents run against a natural range extension. Surprisingly, despite political and scientific focus on NIS hardly any ecological data have been published on this species from Western Australia. We show that B. australis is highly abundant in both seagrass beds (424 ± 29 ind m⁻²) and on unvegetated sand flats (92 ± 22 ind m⁻²) being orders of magnitudes more abundant than any native gastropod in the Swan River. Experiments showed that high resistance to predation and environmental stress potentially explains its success. From our survey data, we calculated that >3.6 billion invasive snails today occupy the Swan River. This large snail populations support other organisms; for example, almost 1 billion macroalgae are found attached to living B. australis and >100 million hermit crabs occupy its empty shells. Given Battilaria’s high abundance, wide distribution, large size, persistent shells that support other organisms and bioturbating behavior, it seems inescapable that this potential invader has impacted the ecosystem functioning of the Swan River. We argue that the search for abundant species of uncertain origin should continue, and that these species generally should be treated with the same interest as high status invaders to mitigate impacts in already invaded systems and to avoid secondary spread into neighboring ecosystems.     

南京下蜀黄土的非均匀性研究

下蜀黄土分布在不同高程的多种地貌单元, 如山顶、阶地、山麓及盆地, 由黄土层和红土层两部分组成, 具有成分和结构差异, 尤其是水稳定性。它是-种非均匀性土层。当其位于地下水位以上或开放性状态时, 土层出现非饱和状态和负孔隙水压力, 产生较大的基质吸力。工程施工中, 可利用这-特性, 但必须控制各种向下净流量, 以抑制土体饱和度增大的趋势。     

结合深度学习的高光谱与多源遥感数据融合分类

高光谱数据具有丰富的光谱特征,但是其空间分辨率相对较低。一些遥感数据具有与高光谱数据互补的优势,例如提供更精细的空间信息的高空间分辨率数据和具有高度信息的激光雷达LiDAR（Light Detection and Ranging ）数据。通过将高光谱数据与多源遥感数据进行融合,可以弥补高光谱数据空间分辨率相对较低,空间特征不够丰富的缺点。近年来,基于深度学习的方法已经在遥感数据分类研究中取得了一定的进展。然而,由于深度网络的特征提取过程是一个自主的过程,往往无法精确的获取最有利于遥感数据分类的特征;同时,深度学习方法具有复杂的网络结构和大量的参数,往往会在分类训练过程中造成参数拟合困难。以上这些因素会导致分类效果不佳。针对这些问题,本文提出了一种将卷积神经网络CNN（Convolutional Neural Network）和纹理特征相结合的多源遥感数据特征级融合分类框架。该方法共3个步骤,首先,对高光谱数据或多源遥感数据提取纹理特征;然后,构造CNN,分别将原始高光谱遥感数据、原始多源遥感数据和第一步中获得的纹理特征作为深度网络的输入进行深度特征提取;最后,将分别提取到的深度特征拼接,并利用Softmax分类器进行分类。为了验证本文提出方法的分类效果,本文在休斯顿和塞特福德矿地区公开数据集上进行实验,并将该分类框架与支持向量机分类方法、像素级融合分类方法和特征级融合分类方法进行对比。由此可以分析得出,本文提出的基于深度学习的融合分类方法可以获得较高的分类精度。