Radon-222 measurements during the Tropoz II campaign and comparison with a global atmospheric transport model

The aim of the ²²²Rn measurements during the airborne campaign TROPOZ II, was first to help in the interpretation of the photochemical studies, and secondly to furnish a data set of ²²²Rn in the troposphere, for validation of atmospheric transport models. In this paper we present the ²²²Rn measurements, and their simulation with a 3-D atmospheric transport model based on observed winds. The ²²²Rn was measured using the active daughters deposit technique with isokinetic aerosol sampling. We have obtained 44 measurements distributed between 65° North and 55° South, from 1 to 11 km height. In 25% of cases, we found relatively high concentrations (> 300 mBq·scm) of ²²²Rn in the high troposphere (>8 km). The results of 3D simulations and the calculations of back-trajectories allow us to find the origins of the high ²²²Rn concentrations. The transport model reproduced most of the observed synoptic variations, but it overestimates the concentrations which implies a vertical transport of excessive velocity.     

The tropospheric distribution of carbon monoxide as observed during the Tropoz II experiment

As part of the TROPOZ II large-scale measurement campaign in January 1991 we deployed a Four Laser Airborne Infra Red (FLAIR) tunable diode laser spectrometer on board a Caravelle 116 research aircraft. We report here in situ CO measurements which were obtained with one of the four channels of the FLAIR instrument at a time resolution of either one or two minutes. The flight route of the TROPOZ II campaign followed the Atlantic coasts of North America, the Pacific and Atlantic coasts of South America and the Atlantic coasts of West Africa and Europe. A total of 48 CO vertical profiles extending from the surface to 10.5 km altitude were obtained. In the meridional direction adjacent profiles were separated by less than 10° latitude. Polewards of 30°S the CO distribution was very homogeneous with a mean mixing ratio of 55 ppbv. Between 30°S and the equator, the CO mixing ratio above 8 km altitude ranged up to 130 ppbv and was 20–60 ppbv higher than in the mid free troposhere. Three day backward trajectories for these CO rich airmasses originated over Amazonia. Earlier trace gas measurements as well as circulation studies suggested that these airmasses were of Northern Hemispheric origin and had been rapidly convected to the upper troposphere over central South America. The influence of biomass burning is clearly apparent from the measurements performed at 10°N on the African side of the Atlantic with CO mixing ratios being 100–300% higher than on the Central American side. CO mixing ratios further north ranged from 80 to 130 ppbv in the free troposphere and increased to 130–150 ppbv at lower altitudes.     

基于GPS-InSAR集成技术地表形变的监测

应用GPS-InSAR集成技术进行地表形变监测是目前极具潜力的研究方向。本文分析了GPS和InSAR两种技术的特点和互补性,提出了GPS-InSAR集成技术数据融合的方案,并通过对国内外研究实例的分析,说明利用GPS-InSAR集成技术监测地表形变是可行的,并具有广阔地应用前景。     

基于矩阵相似度的InSAR图像配准方法研究

图像配准是干涉合成孔径雷达成像的关键技术之一,相关系数是InSAR图像配准的重要指标。本文在分析研究常见相似性度量的基础上,研究了相关系数与矩阵相似度度量之间的关系,研究发现,矩阵相似度度量可涵盖多种图像配准数值指标,且其表达形式精炼,几何意义明确,便于把握图像配准的机理。此外,本文提出了一种加权矩阵相似度度量,并通过仿真数据进行了验证。     

全局收敛LM算法优化的机载激光测深信号提取方法

在机载激光测深系统中,水面与水底回波信号的提取精度是影响系统测深能力的关键因素,然而传统的信号提取方法易受噪声影响,精度低,对测量环境的适应性差。针对上述问题,提出一种全局收敛LM(Levenberg Marquardt)算法优化的机载激光测深信号提取方法。首先通过模糊筛选得到较为可靠的初值;然后利用基于全局收敛LM的波形分解算法对波形进行分析建模,从而获取每个初值点的高斯分解参数;最后通过多条件筛选确定精确的水面、水底回波信号位置。利用实测数据和模拟数据分别进行实验,结果表明提出方法对回波信号具有较高的检测正确率,且稳定性较强。     

一种基于机载LiDAR点云的林间道路提取方法

基于林间道路的形态特征和支持向量机原理(SVM),提出一种从机载LiDAR点云数据中提取林间道路的方法。首先,选取末次回波去噪、栅格化,生成数字表面模型(DSM)和强度信息模型(DIM);然后,通过支持向量机对坡度信息进行分类,得到道路潜在区域;再对道路潜在区域进行强度信息的分类,得到含有少量噪声的初始道路区域;最后,利用形态参数对初始道路区域进行去噪、精化,得出最终道路区域。实验证明,该方法能较好地提取出道路区域。     

机载激光雷达数据中电力线的快速提取

针对当前机载LiDAR技术在电力巡线应用中对电力线数字模型高精度和快速重建的需求,该文提出一种高效的电力线点云分类方法。首先基于局部范围点的高程统计直方图,实现电力线点的快速粗提取;然后运用随机抽样一致性算法剔除残留的电塔点,结合点云高程统计进一步剔除绝缘子点,实现电力线点的精提取;最后利用同一垂直面内电力线点的高程分布特性,实现单根电力线点的快速提取。基于实际输电线路机载LiDAR数据的实验结果表明,该方法可实现电力线点的快速、高精度提取:粗分类后的电力线点中仅含约10%的非电力线点;精分类后约有2%的电力线点被误分为绝缘子点,最终各条电力线点的提取比率平均为98%以上。     

一种分形模型的机载InSAR DEM空洞填补方法

针对如何提高机载InSAR DEM空洞填补精度和DEM对地形细节描述的难题,该文对分形模拟的中点位移插值函数做了改进。首先,提出了动态分形参数提取思路,根据选定的窗口动态提取局部分形参数,使提取的分形参数局部精度显著提高。其次,提出了分形曲面中心内插法,以正方形为图像基元,将两点信息扩展到8点信息来充分显示局部变化的细节信息;考虑到对采样点点数和位置的要求,结合距离加权平均法来弥补周围不够8点的特殊情况。最后,以国产机载X波段全极化InSAR影像提取的DEM为数据源进行了实验。实验结果表明,相比传统的中点分形内插法,该方法不仅能有效提高空洞填补精度,且能更好地描述地形的细节信息。     

相位萃取技术在InSAR小数据集上的应用

相位萃取技术可最小化多基线InSAR处理中干涉像对的去相关噪声,已成为雷达干涉测量研究领域中热点技术之一。本文首先设计并实现了一套专门针对小数据集的基于相位萃取技术的多基线InSAR处理流程。整个流程分为两大模块。第一个模块为称为预处理模块,通过图像配准、地形相位移除等操作生成相位萃取技术所需的零基线SAR图像堆栈。第二个模块首先基于多视像元生成相干矩阵,并且在此基础上完成相位萃取操作。然后,利用小基线集技术的思想在萃取相位上分离出各相位稳定点上的形变情况。为了保证结果可靠性,在这一模块中还加入了大气相位预估与移除的步骤。最后,该流程被应用到了5景拍摄于太原地区的PALSAR图像上。实验结果表明,相位萃取技术在小数据集情况下仍然能够有效提升数据的相干性,这有利于多基线InSAR输出结果密度的提升。     

不同波段垂直指向雷达功率谱密度对比

衰减对回波的影响是雷达探测中的一项重要课题,常用订正方法还有很大提升空间。由于垂直指向雷达观测的功率谱数据包含了非常丰富的微物理及动力信息,该文尝试在功率谱层面对衰减影响机理进行深入讨论,以期改进衰减订正方法。对2014—2016年广东阳江和广东龙门获取的Ka波段毫米波雷达、C波段调频连续波雷达、Ku波段微雨雷达功率谱数据进行对比表明:3部雷达功率谱谱型基本一致,具有一定可靠性。米散射效应反映在功率谱上即经过某一速度后回波强度谱密度的陡降,当降水下落末速度达临界速度时,会对衰减量的计算造成影响。雷达波长越短,受衰减影响越大,衰减对功率谱的影响表现为谱的整体下移,最终导致回波强度小于真值。对于对流云降水,单纯使用经验订正公式获得的结果可能与真实值相差很大,此时需要考虑降水粒子谱分布,在功率谱层面上对订正进一步改善。