基于GPS和MODIS的ENVISAT ASAR数据干涉测量中大气改正方法研究

GPS数据用于改正InSAR中大气延迟误差的方法受GPS站点密度的限制,只利用有限的几个站点所观测到的大气数据来生成干涉图的大气改正图,往往达不到很好的效果.本文研究利用GPS与MODIS数据的联合使用来生成大气改正图,首先用GPS数据对MODIS水汽产品进行分块校准,并且对MODIS水汽数据进行了空间结构函数分析,得到研究区域内水汽场的空间分布规律.然后把这种区域水汽场的空间分布信息结合到Kriging内插法中生成更为合理的水汽图.通过上海地区ENVISAT ASAR数据的实验发现,这种加以改正的GPS和MODIS数据联合改正法不仅可以对长波大气信号有明显的消弱,还能消弱一些短波的大气信号,特别是一些幅度较强的短波信号;经过GPS+MODIS算法改正后,短波信号占优和长波信号占优的两幅差分大气延迟图的整体RMS分别降低了32.74%和38.82%,去除幅度较大.与GPS+ATM(大气传输模型)算法比较,我们发现,在上海地区有限的数据条件下（即研究区域内只有6个GPS点）,GPS+MODIS法在大气去除效果或者说大气信号重现能力方面优于GPS+ATM算法.GPS+MODIS算法在捕获短波大气信号方面要比GPS+ATM更有优势,因此也可以改正短波大气误差.     

利用MERIS水汽数据改正ASAR干涉图中的大气影响

大气对流层对雷达信号的传播延迟是制约重复轨道InSAR高精度测量应用的重要因素之一.本文描述了MERIS水汽数据用于ASAR干涉图大气改正的方法;并以美国南加州地区为例,选取4对ENVISAT ASAR数据进行了大气改正的研究.结果显示对这4幅干涉图,经过MERIS水汽数据改正后InSAR与GPS差异的RMS分别〖JP2〗降低了41.7%,65.2%,19.3%和39.4%.平均改善程度达41.4%.更重要的是,经过MERIS水汽改正后,从2005~2007年〖JP〗干涉图和2004~2007年干涉图中,能清楚地识别出三处形变最明显的区域：Long Beach-Santa Ana 盆地、Pomona-Ontario和San Bernardino,其形变速率从-8 mm/a到-28 mm/a,大部分在-20 mm/a左右,与这些地区2003年以前的历史形变速率基本一致.因此,采用无云条件下的MERIS水蒸汽数据改正同步获取的ASAR干涉图,可以显著地降低大气水汽对干涉图相位的影响,从而更真实地反映地表形变等地球物理信号.     

一种InSAR大气相位建模与估计方法

为了削弱大气延迟对干涉结果的影响以提高InSAR的测量能力,本文在InSAR大气相位特征分析的基础上,研究了一种新的InSAR大气相位建模与估计方法.首先采用稳健估计确定大气垂直分层部分的模型参数,然后利用基于Matern模型的Kriging插值估计大气紊流部分,最后应用估计的大气垂直分层和紊流资料改正InSAR测量结果.利用覆盖河南义马地区的ASAR数据对本文提出的方法进行了验证,结果表明去除大气影响后,InSAR重建的DEM与参考DEM的高程差异的均方误差由19.5m降至5.3m,精度提高了约72%.同时,改正后的干涉图更合理地揭示了义马矿区的沉降漏斗情况,进一步验证了本文方法的有效性.     

基于MERIS水汽数据的ASAR干涉图改正

水汽延迟是InSAR技术的重要误差源,已成为高精度差分干涉测量的主要限制.本文比较分析了GPS与MERIS可降水量反演结果以及水汽变化量探测结果,研究发现两者探测到的水汽含量和水汽变化量均存在较好一致性.在此基础上,以南加州地区为例验证了差分MERIS PWV用于同步ASAR干涉图水汽改正的可行性,经过水汽改正干涉图质量提高了约23%,并进一步对比了InSAR形变监测结果与GPS探测到的形变,比较结果表明,消除水汽影响后干涉图形变结果可靠性提高了19%.     

InSAR大气误差改正及其在活动断层形变监测中的应用

为了获取青藏高原东北缘老虎山断裂带精确的震间形变速率场,通过对短时间基线干涉图大气改正效果的评价,从3种外部大气数据(MERIS, ERA-Ⅰ, WRF)中确定出最优的大气改正方法,用于长时间基线干涉图中的大气信号改正;然后利用层叠法(stacking)累积平均经大气和轨道改正后的干涉图,获取了研究区的震间形变速率场.结果显示:海原断裂系统区域内,MERIS和ERA-Ⅰ的大气改正效果优于WRF;MERIS和ERA-Ⅰ的改正结果给出了相似的形变速率场,断层两盘相对形变速率为视线向2.5 mm/a,转换成平行于断层方向为6.5 mm/a,与GPS结果一致;在近断层5 km的范围内,出现了较大的形变梯度,揭示了浅层蠕滑的存在.      

基于全球气象再分析资料的InSAR对流层延迟改正研究

气压、温度和水汽含量等大气物理参数的时空变化导致的对流层延迟是制约合成孔径雷达干涉测量(Interferometric Synthetic Aperture Radar,InSAR)高精度应用的重要因素之一.最新研究显示气象再分析资料在补偿对流层延迟影响方面具有巨大的应用潜力,这促使我们对其有效性和鲁棒性做进一步的研究和探索.本文首先推导了利用气象再分析资料对InSAR进行对流层延迟校正的算法;然后以美国南加州地区的ENVISAT ASAR数据为例,分析了基于两种气象再分析资料(ERA-Interim和North American Regional Reanalysis,NARR)校正InSAR对流层延迟改正的效果;通过与MERIS水汽延迟改正结果比较,验证了该方法的有效性.实验结果表明:(1)不能简单忽略干延迟,可通过气象再分析资料进行有效估计;(2)通过与MERIS水汽产品获得的对流层延迟比较发现,气象再分析资料能够取得接近于MERIS的改善效果;(3)对ERA-Interim和NARR两种气象再分析资料而言,虽然后者具有更高的时间和空间分辨率,但在改正InSAR对流层延迟方面并没有表现出比前者更明显的优势;(4)气象再分析资料可以很好地估计与地形强相关的垂直分层延迟,但对于小尺度的湍流混合延迟的捕捉能力有限.综合分析认为,气象再分析资料的优势在于其数据可随时获得、免费和全球覆盖,它可以显著减弱大尺度的垂直分层延迟对干涉图相位的影响,从而有助于InSAR获取更真实可靠的地形高程和地表形变信息.     

高光谱遥感数据的改正暗目标大气校正方法研究

高光谱遥感数据常用的大气校正方法均侧重于去除水汽及其他吸收气体的影响, 主要研究了从高光谱影像同时去除气溶胶与水汽影响的方法. 由于高光谱遥感数据波段众多, 常规暗目标方法一直难以适用于高光谱遥感数据的大气校正. 通过选取小麦作为新的暗目标对象, 着重讨论了使用多波段线性回归与插值的方法对常规暗目标方法进行改正使之充分利用高光谱的众多波段特性, 从而把改进的暗目标方法扩展应用于高光谱遥感数据的大气校正. 为了同时去除大气中水汽的影响, 大气校正全过程采用了循环迭代的算法. 以山东济宁地区EO-1卫星搭载的Hyperion高光谱数据为应用实例, 通过使用MODTRAN建立的查找表直接从影像估算出气溶胶与水汽含量, 实现了对该数据的大气校正. 大气校正的结果表明, 改正暗目标大气校正算法可以有效地对高光谱遥感数据进行大气校正.     

高级时序InSAR地面形变监测及地震同震震后形变反演

差分干涉测量技术(differential interferometry technique,D-InSAR)是在InSAR技术基础上发展起来的,主要应用于测量地表微小形变。该技术在监测地震形变、火山活动、冰川运动、城市地面沉降以及山体滑坡等领域得到广泛应用,但实际研究中存在很多局限性,例如轨道参数误差、地形数据误差、大气延迟误差、时空去相关引起的相位解缠误差以及系统噪音误差等。这些因素成为InSAR高精度形变探测应用的主要限制。InSAR时序分析方法是目前解决常规D-InSAR处理过程中各种精度制约因素的主要方法之一。本文以InSAR时序分析中几个关键技术问题为研究目标,开展InSAR时序关键技术研究。本文取得的创新点如下:(1)在传统相干点提取方法的基础上,提出了一种新的提取方法:满秩相干点。该方法将干涉冗余网络图中节点和边结合起来构建满秩矩阵,并依此作为相干点提取指标。实验结果表明,满秩相干点提取法不仅能有效提高相干点的空间分布密度,而且能保留相干点的最优测量精度。(2)将大气延迟相位分成3种分量,即长波长大气相位、短波长大气相位以及地形相关的大气相位。在此基础上,利用网络法分别对3种类型大气延迟相位进行校正。网络法可以估算出每个时刻的大气延迟误差,再重构每一幅干涉图的大气延迟误差。实验结果分析表明,采用网络法能精确模拟出每个时刻大气延迟误差和每幅干涉图中的大气延迟相位。(3)离散点相位解缠误差是时序分析过程中主要误差源之一,直接影响后续时序分析结果的精度。本文提出一种利用网络闭合环残差方法,对离散相干点相位解缠误差进行检测和校正。该方法可以有效识别出离散相干点解缠相位中相位跳变;同时,本文也介绍了离散相干点解缠相位跳变校正的方法。(4)基于相干矩阵满秩条件,提出一种新的InSAR时序反演策略。利用干涉冗余网络图中各条边和节点之间的关系,采用最小二乘方法解算各个时刻点的形变特征,以有效解决传统InSAR时序反演过程中的秩亏问题。通过Bam地震震后时序形变场提取,在羊八井电站地热开采地下水引起地面沉降监测以及北京市地面沉降监测等几种典型应用中,验证了该方法的有效性和精确性。在关键技术突破的基础上,本文利用提出的新方法,获取了2008年10月6日西藏当雄MW6.3地震同震形变和震后地表形变时序,研究发震断层参数和震后运动过程的动力学机制。利用同震形变反演得到断层几何参数及滑动分布模型为先验知识,以InSAR时序方法获得震后形变时序资料为约束,分别利用震后余滑模型和震后黏弹性松弛模型,研究当雄地震震后断层的运动过程,同时对青藏高原南部中低地壳或者上地幔的黏性系数进行估计,并给出了最优拟合解。     

GPS大气掩星技术在全球气候变化研究中的应用

人类活动引起全球变暖,衡量全球气候变化的指标有陆地、大气和海洋温度,水汽含量等等.研究对流层底层大气温度和水汽含量变化的传统方法是用数值天气预报模型和微波声纳,尚未实现用全球均匀覆盖的数据来做精确的定量研究.和GNSS系列卫星计划比较,最近发射的COSMIC卫星气象探测数据的空间、时间以及垂直分辨率都大大提高.采用COSMIC数据可以改进和量化南极洲的大气压力模型,并综合GNSS系列卫星测量的水汽和温度剖面研究全球气候变化.用一维协方差算法估计南极洲及附近海洋的大气压、温度和湿度剖面.把COSMIC卫星密集测量期间演算得到的大气折射率和GNSS系列卫星的结果进行比较.再和独立测量数据进行比较,包括南极洲自动气象观测站资料,数值天气预报模型资料,多种测高卫星水汽资料和海洋表面温度资料以及区域GPS水汽图.上述工作将改进发展中的气象遥感技术并应用于天气预报和空间天气预报及全球气候变化研究.     

基于星载高精度GPS观测数据的大气密度反演

星载高精度GPS观测数据可提供卫星速度和位置信息,而卫星的运行轨迹又与所处位置的大气密度紧密相关,因此可通过求解大气阻力微分方程,由高精度GPS观测数据反演出卫星运行轨迹上的热层大气密度.本文从星载高精度GPS观测数据出发,给出大气密度的反演方法,以及平均平动参数nM、反弹道系数B两个重要参数的解算过程,并以天宫一号为例,给出反演结果与天宫一号观测数据的比对.结果表明,反演结果与观测值符合很好,两者的均方差在2012年1月1日、2月24日分别为8.6%和8.4%,说明利用星载GPS观测数据反演大气密度是有效、可行的,可成为今后获取高精度大气密度的一种方法.     

A Quantitative Assessment of DInSAR Measurements of Interseismic Deformation: The Southern San Andreas Fault Case Study

We investigate the capabilities and limitations of the Differential Interferometric Synthetic Aperture Radar (DInSAR) techniques, in particular of the Small BAseline Subset (SBAS) approach, to measure surface deformation in active seismogenetic areas. The DInSAR analysis of low-amplitude, long-wavelength deformation, such as that due to interseismic strain accumulation, is limited by intrinsic trade-offs between deformation signals and orbital uncertainties of SAR platforms in their contributions to the interferometric phases, the latter being typically well approximated by phase ramps. Such trade-offs can be substantially reduced by employing auxiliary measurements of the long-wavelength velocity field. We use continuous Global Positioning System (GPS) measurements from a properly distributed set of stations to perform a pre-filtering operation of the available DInSAR interferograms. In particular, the GPS measurements are used to estimate the secular velocity signal, approximated by a spatial ramp within the azimuth-range radar imaging plane; the phase ramps derived from the GPS data are then subtracted from the available set of DInSAR interferograms. This pre-filtering step allows us to compensate for the major component of the long-wavelength range change that, within the SBAS procedure, might be wrongly interpreted and filtered out as orbital phase ramps. With this correction, the final results are obtained by simply adding the pre-filtered long-wavelength deformation signal to the SBAS retrieved time series. The proposed approach has been applied to a set of ERS-1/2 SAR data acquired during the 1992–2006 time interval over a 200?×?200?km area around the Coachella Valley section of the San Andreas Fault in Southern California, USA. We present results of the comparison between the SBAS and the Line Of Sight (LOS)—projected GPS time series of the USGC/PBO network, as well as the mean LOS velocity fields derived using SBAS, GPS and stacking techniques. Our analysis demonstrates the effectiveness of the presented approach and provides a quantitative assessment of the accuracy of DInSAR measurements of interseismic deformation in a tectonically active area.     

基于PSInSAR技术的海原断裂带地壳形变初步研究

常规差分干涉测量(DInSAR)受时间、空间失相干的严重制约和和大气延迟等相位误差的影响,难以实现对长期累积微小地壳形变场的有效探测.PSInSAR技术克服了常规DInSAR的局限性,能够高精度监测微小地壳形变.本文首先介绍了PSInSAR技术的算法模型和处理方法.该方法通过二维线性相位模型,对时序干涉图象上相干点目标...     

ESTIMATING PRESENT SLIP RATE OF THE FAULTS IN THE WEIHE GRABEN USING ENVISAT ASAR DATA

Most great(M≥8)earthquakes during modern times have occurred in interplate regions or major continental collision zones, such as Sumatra, the Japanese island arc or the San Andreas fault zone. Continental faults slip at a much lower rate than boundary faults, but they also have the potential of generating large earthquakes. For example, the 2008 Wenchuan earthquake with a magnitude of 7.9, the slip rate of seismic fault is less than 3mm/a. They also have the potential to be significantly deadlier than those on plate boundaries because of the long repeat times and lack of preparedness. The January 23rd 1556 Huaxian earthquake in Shaanxi Province, central China, is the deadliest in history with an estimated death toll of ~830 000 from building collapse, land-sliding, famine, and disease. The earthquake occurred in the graben of the Weihe River.
The Weihe Graben in Shaanxi Province has recorded multiple earthquakes in history, whereas most active faults within the graben have a low slip rate over geological times (~1mm/a). The slip rate of faults is an important parameter for assessing the risk of earthquakes and the interval between major earthquake recurrences. In order to obtain the quantitative information of faults slip rate, traditional geological methods or geodetic observation techniques can be used. Interferometric synthetic aperture radar(InSAR), as a modern geodetic observation technology, has the characteristics of all-weather and day-and-night imaging capability, wide spatial coverage, fine resolution, and high measurement accuracy. InSAR offers the potential to measure interseismic slip rates on faults at a resolution of millimetres per year. In this study, we use InSAR data to analyze the present deformation of the Kouzhen-Guanshan, Weihe and North Qinling faults in the central part of the graben.
We collected 32 European Space Agency(ESA's)Envisat ASAR images from descending track 161 between 2003 and 2010, and processed them using ROI_PAC. The precise orbit determination from the Delft Institute for Earth Oriented Space Research(DEOS)was applied to correct for orbital effects. The topographic contribution was simulated and removed using the 90m resolution Shuttle Radar Topography Mission(SRTM)Digital Elevation Model(DEM)from CGIAR-SCI. Each interferogram was downsampled to 64 looks in the range direction (1 280m). Before phase unwrapping, a weighted power spectrum filter was applied to improve the signal-to-noise ratio. The branch-cut method was used for phase unwrapping. Phase unwrapping errors were checked by summing around a closed loop. All the major unwrapping errors were identified and corrected manually. We obtained a total of 98 interferograms with a spatial baseline of smaller than 300m, and selected 33 interferograms whose coherence is well preserved for time-series analysis. The time-series analysis was implemented using the π-RATE software package. It uses the geocoded interferograms from ROI_PAC to create a minimum spanning tree(MST)network, from which the orbital and topographically-correlated atmospheric errors are estimated. The MST network connects all epochs with the most coherent interferograms,including no closed loops of interferograms. The network approach is able to improve the estimation of orbital error by ~9% compared to the independent interferograms approach. The orbital errors are empirically modelled as planar or quadratic ramps. The topographically-correlated atmospheric correction was applied to each interferogram after having corrected for the orbital errors. Following creating a minimum spanning tree network, correcting for orbital and topographically-correlated atmospheric errors, and calculating the covariance matrix, we obtained the 7-year average slip rate of the faults that we are focused on.
Our results show that the faults across the Weihe graben all have a small slip rate of less than 2mm/a. The Kouzhen-Guanshan Fault does not show any evident deformation signal. The Weihe Fault seems to show 1mm/a normal faulting in the satellite line-of-sight direction. In addition, we find ~10mm/a surface subsidence of the Xi'an City between 2003 and 2010. We use the stable Ordos block as a reference to assess the accuracy of our InSAR time-series analysis. Assuming the Ordos block has no internal deformation, we calculated the error of the InSAR rate map to be (-0.1±1)mm/a, indicating that our result is reliable. This paper presents a preliminary result of the present deformation of the Weihe Graben. InSAR is a powerful technique for monitoring active faults on a timescale of tens of years, and can be used for seismic hazard assessment in the future.     

COSMIC大气掩星开环数据反演方法

COSMIC星座GPS无线电掩星探测利用GPS开环接收技术提高低层大气观测数据的质量和对上升掩星事件的跟踪能力.开环掩星观测数据受到GPS导航数据调制的影响,在其数据后处理中必须消除该影响以获得高质量的科学反演结果.利用GPS导航数据调制码数据和利用开环数据本身内在的关系等两种方法可以消除该影响.将上述方法应用于COSMIC的掩星事件个例反演,获得了修复的大气附加相位数据;并利用几何光学近似反演方法和全谱反演方法,获得了射线弯曲角. 全谱反演方法获得的弯曲角及其温度反演结果与COSMIC数据中心的结果一致,说明我们的算法是有效的.     

Atmospheric effects on repeat-pass InSAR measurements over Shanghai region

It is well known that interferometric synthetic aperture radar (InSAR) measurements are affected by the atmosphere especially the troposphere. We use Shanghai as an example to analyze the properties of the effects based on a number of SAR interferograms over the region. First, Radon transform is used to examine the isotropic property of the effects. The results show that the effects in all the interferograms are anisotropic although they exhibit different patterns. Two different methods, the Jarque-Bera and the Hinich tests, are then applied to test the Gaussianity of atmospheric effects. Both of the tests reveal that the atmospheric effects are non-Gaussian. Finally, spectral analysis of the atmospheric effects is carried out and the results show that the power spectra of the atmospheric signatures in all the interferograms studied follow approximately the power law distribution. The scaling exponents estimated show that the atmospheric signatures in all the interferograms are non-stationary but with stationary increments. It is estimated based on the estimated scaling exponents that external data with spatial resolutions as high as about 0.3 km, are required to determine and correct for over 90% of the atmospheric effects in the SAR interferograms.     

大气边界层湍流的动力非平稳性的验证

首次用验证时间序列中是否存在动力非平稳性的一种简单图示方法——space time index法来分析大气边界层湍流的动力平稳性特征.本文以取自淮河流域和威斯康星森林下垫面条件下的三维高精度风速和温度、湿度湍流脉动资料对大气边界层湍流的平稳性特征进行了分析.结果表明space time index方法能有效地检验大气边界层湍流信号中是否存在动力平稳性.另外,均匀下垫面条件（水稻田）及复杂下垫面条件（森林）下的大气边界层湍流信号中几乎都存在动力非平稳性,大气湍流动力学非平稳性可能是边界层湍流信号相当普遍具有的一种特性.大气边界层湍流中的间歇性和相干结构使得其非平稳性图形的特征不同于一般时间序列非平稳性图形的“V”型特征;森林下垫面条件下的湍流信号比相对均匀下垫面（水稻田）下的湍流信号更有组织性,相干结构更强.