利用GIS技术和RS技术减灾防灾

减灾防灾是一项需要多学科知识融会贯通的长期工程。灾难的发生是由多种因素造成的,在分析管理的时候,需要大量的数据源,因而,分析过程相当地耗时耗费。近年,由于能够高效提供详细、精确的相关数据．GIS技术和RS技术在灾难管理方面应用广泛、发展迅速,遥感技术提供灾情数据．GIS技术通过模型将各种数据有机地整合起来,包括空间数据、非空间数据、属性数据,使它们在灾难管理的各个阶段发挥不同的作用。     

Distribution and significance of novel low molecular weight sterenes in an immature evaporitic sediment from the Jinxian Sag,North China

An unusual series of C₂₂–C₂₇ monounsaturated sterenes and C₂₄–C₃₀ tetracyclic terpanes (17,21-secohopanes) were detected in relatively high concentrations in an immature evaporitic marl sediment of the Jinxian Sag, Bohai Bay Basin, North China. The site of unsaturation in these novel sterenes is assigned tentatively to the D ring on the basis of mass spectral interpretation, which also distinguishes them from reported unsaturated sterenes. Other hydrocarbon biomarker or stable isotope characteristics are indicative of microbial (e.g. methyl hopanes), phytoplankton or higher plant (depleted δ¹³C values of isoprenoids and hopanes) inputs and an anoxic carbonate depositional environment (hexacyclic hopanes; tetracyclic terpanes). The hydrocarbon composition showed no obvious biodegradation and the relatively high concentration of unsaturated terpenoids (e.g. gammacerene) and low values of other established maturity parameters (T_s/T_m = 0.23; R_o = 0.44%; T_max = 417 °C), are consistent with sediments of low maturity. The novel, low molecular weight sterenes and the tetracyclic terpanes may be early diagenetic products of microbial sources in a carbonate environment.     

柴达木盆地东北部新近纪构造旋转及其意义

青藏高原东北缘构造变形的研究是认识高原隆起过程、机制和印度—欧亚板块碰撞远程效应的重要途径。柴达木盆地是印度－欧亚板块碰撞后南北向挤压应力为动力背景的高原东北部内陆盆地,沉积物主要来自于周边山地,完整的保存了新生代以来高原隆升的详细记录。通过柴达木盆地东北部瑙格剖面精细古地磁及构造旋转研究发现,20.1～15.1Ma以及15.1～8.2Ma柴达木盆地分别发生了9.7°±7.4°和6.4°±4.4°的顺时针旋转,约8.2Ma后,柴达木盆地东北部瑙格地区发生了16°±7.5°的逆时针快速旋转。通过分析认为,前两次的顺时针构造旋转事件可能与阿尔金断裂的左旋走滑有关。而约82Ma以来的逆时针旋转事件属于柴达木盆地东北部瑙格地区的局部旋转,可能与温泉断裂的右旋走滑有关,说明青藏高原东北部在昆仑山、阿尔金山和祁连山三条巨型断裂系左旋相对运动的宏观控制下形成的NNW向温泉右旋走滑断裂开始走滑的年代为约8Ma。     

Transport of waste heat from a nuclear power plant into coastal water

A numerical and experimental modeling is presented for studying the transport of waste heat from a nuclear power plant into coastal water by using a full-field physical model with scale distortion, a local physical model with normal scale and a depth-averaged k− turbulence model with a modified second-order upwind scheme. Field investigations are also used to provide the calibration and validation of the modeling. A case study simulating the turbulent tidal flow and waste heat transport in the coastal water near Daya Bay Nuclear Power Plant in Southern China was conducted. The experimental result of the case study shows that the water temperature in coastal water was a little oversimulated near the surface and was a little undersimulated near the bottom of heated-water layer by the full-field physical model. The numerical study shows that the depth-averaged k− turbulence model presented a satisfied prediction of turbulent tidal flow and far-field temperature distribution in coastal water, although the near-field stratification due to the heated water effluent was not accounted for. The result of the effect of scale distortion on physical model shows that a full-field physical model with a scale distortion of 10 produced a satisfied result of temperature distribution in the present case study.     

基于分类回归树分析的遥感影像土地利用／覆被分类研究

以专家知识和经验为基础,综合影像光谱信息和其他辅助信息进行分类的基于知识的遥感影像解译方法,是提高遥感影像分类精度,实现自动解译的有效途径之一。然而,知识的获取一直是其得以广泛应用的“瓶颈”问题。以江苏省江宁试验区土地利用/覆被分类为例,利用分类回归树分析(CART)从训练样本数据集中发现分类规则,集成遥感影像的光谱特征、纹理特征和空间分布特征进行分类实验,并与传统的监督分类和逻辑通道分类方法进行比较。结果表明,基于CART的分类方法的精度基本在80%以上,与另两种方法相比,有了较大的提高,而且该算法复杂性低,效率高。由此说明,利用CART算法构建决策树获取的分类规则是合理的。它可以快速、有效地获取大量分类规则,是促进基于知识的遥感影像分类方法在土地利用/覆被分类中广泛应用的一项有效手段。     

典型遥感数据分类方法的适用性分析——以遥感图像场景分类为例

目前普遍采用的分类器通常都是针对单一或小量任务而设计的,在小数据量的处理中能取得比较满意的结果。但对于海量遥感数据的处理,其在处理时效和分类精度方面还有待研究。本文以遥感图像场景分类任务为例,着重对遥感数据分类问题中几种典型分类方法的适用性进行比较研究,包括K近邻(KNN)、随机森林(RF),支持向量机(SVM)和稀疏表达分类器(SRC)等。分别从参数敏感性,训练样本数据量,待分类样本数据量和样本特征维数对分类器性能的影响等几个方面进行比较分析。实验结果表明:(1)KNN,RF和L0-SRC方法相比RBF-SVM,Linear-SVM和L1-SRC,受参数影响的程度更弱;(2)待分类样本固定的情况下,随着训练样本数目的增加,SRC类型分类方法的分类性能最佳,SVM类型方法次之,然后是RF和KNN,在总体分类时间上呈现出L0-SRCL1-SRCRFRBF-SVM/Linear-SVMKNN/L0-SRC-Batch的趋势;(3)训练样本固定的情况下,所有分类方法的分类精度几乎都不受待分类样本数目变化的影响,RBF-SVM方法性能最佳,其次是L1-SRC,然后是Linear-SVM,最后是RF和L0-SRC/L0-SRC-Batch,在总体分类时间上,L1-SRC和L0-SRC相比其他分类方法最为耗时;(4)样本特征维数的变化不仅影响分类器的运行效率,同时也影响其分类精度,其中SRC和KNN分类器器无需较高的特征维数即可获得较好的分类结果,SVM对高维特征具有较强的包容性和学习能力,RF分类器对特征维数增加则表现得并不敏感,特征维数的增加并不能对其分类精度的提升带来更多的贡献。总的来说,在大数据量的遥感数据分类任务中,现有分类方法具有良好的适用性,但是对于分类器的选择应当基于各自的特点和优势,结合实际应用的特点进行权衡和选择,选择参数敏感性较小,分类总体时间消耗低但分类精度相对较高的分类方法。     

Feature Construction and Identification of Convective Wind from Doppler Radar Data

Convective wind is one of the common types of severe convective weather. Identification and Forecasting of con- vective wind are essential. In this paper, five ...     

Investigation of the correlation between morphology and luminosity for two classes of main galaxies

Using an apparent-magnitude limited Main galaxy sample of the Sloan Digital Sky Survey Data Release 7(SDSS DR7), we investigate the correlation between morphologies and luminosity for the Main galaxy sample. Our Main galaxy sample is divided into two classes: Main galaxies only with TARGET_GALAXY flag (bestPrimtarget = 64), and ones also with other flags. It is found that for the second class Main galaxies, the early-type proportion monotonously increases with increasing luminosity nearly in the whole luminosity region. But for the first class Main galaxies, the early-type proportion increases with increasing luminosity only within a certain luminosity region (−22.2 < M _r  < −19.8). In the high luminosity region (M _r  < −22.2), the early-type proportion of the first class Main galaxies even decreases dramatically with increasing luminosity. We also analyze the correlation between morphologies and luminosity of galaxies around the peak of the redshift distribution ( 0.07 ≤ z ≤ 0.08 ). In such a narrow redshift region, we still observe strong correlation between morphologies and luminosity, which shows that this correlation is fundamental.     

瞬变电磁扩散场到虚拟波场的精细积分变换算法

现有的瞬变电磁解释与反演方法主要对电阻率参数进行解释,较难给出准确的地质构造信息.为了能够较多获得地质构造信息,可将瞬变电磁扩散场信号转换为虚拟波场信号,提高瞬变电磁法的分辨率.本文采用精细积分法,实现了瞬变电磁扩散场到虚拟波场的转换.将一个高度病态的线性方程组转换成求积分的过程,大大降低了解决病态问题的难度.积分步长以2的指数增加,在保证精度的前提下,具有极高的计算效率;设置合理的终止迭代条件,进一步提高本文方法的适用性.对典型地电模型的计算,表明本文方法具有较高的精度;减小两个波峰之间的距离,证明了本文方法具有较好的分辨能力;对含有噪声信号的扩散场进行波场反变换,证明了本文方法具有较好的抗噪性.最后计算三维模型和实测数据的虚拟波场,可以很好地反映出三维地质体的界面信息,证明本文方法稳定、可靠,可以对地下地质目标体的界面进行有效识别.本文方法相比预条件正则化共轭梯度法(PRCG),在计算效率方面提升了 4倍左右;本文方法的相对误差在5％以内,而PRCG的最大相对误差可达40％～50％,精度最高可提升10倍;在分辨率与稳定性方面,也表现出较好的效果.