D-InSAR技术在盘锦地区地面沉降监测中的应用研究
发布时间：2018-08-31 11:55
【摘要】：地面沉降(Land subsidence)是由于自然因素和人为因素造成的地表垂直向变形现象。我国很多城市和地区都受到地面沉降的影响。地面沉降直接影响了社会稳定和区域经济的持续发展。 合成孔径雷达干涉差分测量(D-InSAR)技术是近几年发展起来的新技术，，主要用于地面沉降、山体滑坡、冰川移动、火山活动、地震等地表微小形变的监测。其精度可以达到厘米级。与GPS和水准测量相比，D-InSAR技术具有覆盖范围大、监测成本低、获取数据快、监测难以监测的区域、监测精度高的优势。 本文利用合成孔径雷达差分干涉测量(D-InSAR)技术，选取2007-2011年的ALOS-PALSAR数据6幅（景），结合STRM DEM数据，通过影像配准、干涉图滤波、去平地效应、相位解缠、去地形相位和地理编码等技术流程，对盘锦地区的地面沉降开展了研究。 盘锦地区存在东郭苇厂（A）、欢喜岭（B）、西八千乡（C）、盘锦市（D）4个地面沉降区。其中东郭苇厂（A）沉降区面积4.22km2，椭圆形沉降区长轴方向为北东—南西向，2007-2011年累计沉降量为-746mm，平均沉降速率为228mm/a；欢喜岭（B）沉降区形态近似圆形，面积0.94km2，2007-2011年累计沉降量为-263mm，平均沉降速率为86mm/a；西八千乡（C）沉降区面积4.38km2，椭圆形沉降区长轴方向为北东—南西向，2007-2011年累计沉降量为-414mm，平均沉降速率为124mm/a；盘锦市（D）沉降区面积较小，多为点状分布，2007-2011年累计沉降量为-70mm，平均沉降速率为23mm/a。 通过地面沉降的时间序列分析，东郭苇厂（A）沉降面积有增大的趋势，但沉降速率有放缓的趋势；欢喜岭（B）和西八千乡（C）沉降速率波状起伏并有逐渐变缓趋势，欢喜岭（B）沉降面积逐渐变小，西八千乡（C）沉降面积有增大趋势；盘锦市（D）沉降速率波状起伏。 采用枝切树法和最小费用流法对相位图解缠进行了对比研究。第一种差分方法直接对干涉相位解缠，通过最小二乘法确定模拟地形相位的缩放因子，从而改进模拟的地形相位。第二种差分方法不需要对干涉相位（不是差分相位）解缠，即使干涉相位不能成功解缠，也能计算出合理结果。两种不同解缠方法得到的计算结果基本一致。在东郭苇厂（A）、欢喜岭（B）、西八千乡（C）和盘锦市（D）4个沉降区，用枝切树法解缠得到的最大形变量分别为-176mm、-88mm、-99mm和-28mm，用最小费用流方法解缠得到的最大形变量分别为-169mm、-78mm、-105mm和-16mm，最小费用流方法解缠比枝切树法解缠得到的计算结果普遍偏小。枝切树解缠法适合平原地区解缠计算，且解缠速度较快；最小费用流法考虑全局最优，适合山区和平原地区解缠。
[Abstract]:Land subsidence (Land subsidence) is a vertical deformation phenomenon caused by natural and human factors. Many cities and regions in China are affected by land subsidence. Land subsidence directly affects social stability and the sustainable development of regional economy. Synthetic Aperture Radar Interferometric differential Measurement (D-InSAR) is a new technique developed in recent years. It is mainly used in the monitoring of ground subsidence, landslide, glacier movement, volcanic activity, earthquake and so on. Its precision can reach centimeter level. Compared with GPS and leveling, D-InSAR technology has the advantages of wide coverage, low monitoring cost, fast data acquisition, difficult monitoring and high monitoring accuracy. In this paper, ALOS-PALSAR data from 2007-2011 are selected by using synthetic Aperture Radar differential Interferometry (D-InSAR) technique. Combined with STRM DEM data, image registration, interferogram filtering, levelling effect, phase unwrapping are used. The land subsidence in Panjin area is studied in the process of detopographic phase and geographic coding. In Panjin area, there are four ground subsidence areas in (A), Huanxiling of Dongguo Reed Plant, (C), Panjin City, (C), Xieqianxiang, (B),. The area of (A) settlement area in Dongguo Reed Plant is 4.22 km ~ 2, the long axis of elliptical settlement area is -746mm from 2007 to 2011, the average subsidence rate is 228mm / a, and the shape of Xiling (B) settlement area is approximately circular. The accumulative subsidence area of 0.94km2 / s in 2007-2011 is -263mm, the average subsidence rate is 86mm / a, the area of (C) subsidence area in Xibaqianxiang is 4.38km2, the long axis direction of elliptical subsidence zone is -414mm in 2007-2011, the average subsidence rate is 124mm / a; In Panjin City, the area of (D) subsidence area is small, and the accumulative subsidence amount is -70 mm from 2007 to 2011, and the average subsidence rate is 23 mm / a. Based on the time series analysis of land subsidence, the settlement area of (A) in Dongguo Reed Plant has a tendency to increase, but the settlement rate has a tendency to slow down, while the settling rate of (B) in Huanxiling and (C) in Xibaqianxiang fluctuates wave-like and gradually slows down. The settlement area of (B) in Huanxiling gradually becomes smaller, and the subsidence area of (C) in Xibaqianxiang has an increasing trend, while the settling rate of (D) in Panjin fluctuates in waves. The twisted-cut tree method and the least cost flow method are used to study the phase diagram wrapping. The first difference method unwraps the interferometric phase directly and determines the scaling factor of the simulated terrain phase by the least square method so as to improve the terrain phase of the simulation. The second method does not need to unwrap the interference phase (not the differential phase), even if the interference phase can not be unwrapped successfully, the reasonable results can be calculated. The results obtained by two different unwrapping methods are basically consistent. In the four subsidence areas of (A), Huanxiling (B), Xibaqianxiang (C) and Panjin (D) in Dongguo Reed Factory, The maximum shape variables obtained by twisted-tangential tree method are -176mm-88mm-99mm and -28mm, respectively, and the maximum shape variables by the least cost flow method are -169mm / -78mm / -105mm and -16mm respectively. The results obtained by the least cost flow method are generally smaller than those obtained by the twisted-tangent tree method. The branch cutting tree unwrapping method is suitable for the calculation of unwrapping in plain area, and the minimum cost flow method is suitable for the global optimization, and is suitable for the unwrapping in mountainous and plain areas.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P642.26
本文编号：2214910
[Abstract]:Land subsidence (Land subsidence) is a vertical deformation phenomenon caused by natural and human factors. Many cities and regions in China are affected by land subsidence. Land subsidence directly affects social stability and the sustainable development of regional economy. Synthetic Aperture Radar Interferometric differential Measurement (D-InSAR) is a new technique developed in recent years. It is mainly used in the monitoring of ground subsidence, landslide, glacier movement, volcanic activity, earthquake and so on. Its precision can reach centimeter level. Compared with GPS and leveling, D-InSAR technology has the advantages of wide coverage, low monitoring cost, fast data acquisition, difficult monitoring and high monitoring accuracy. In this paper, ALOS-PALSAR data from 2007-2011 are selected by using synthetic Aperture Radar differential Interferometry (D-InSAR) technique. Combined with STRM DEM data, image registration, interferogram filtering, levelling effect, phase unwrapping are used. The land subsidence in Panjin area is studied in the process of detopographic phase and geographic coding. In Panjin area, there are four ground subsidence areas in (A), Huanxiling of Dongguo Reed Plant, (C), Panjin City, (C), Xieqianxiang, (B),. The area of (A) settlement area in Dongguo Reed Plant is 4.22 km ~ 2, the long axis of elliptical settlement area is -746mm from 2007 to 2011, the average subsidence rate is 228mm / a, and the shape of Xiling (B) settlement area is approximately circular. The accumulative subsidence area of 0.94km2 / s in 2007-2011 is -263mm, the average subsidence rate is 86mm / a, the area of (C) subsidence area in Xibaqianxiang is 4.38km2, the long axis direction of elliptical subsidence zone is -414mm in 2007-2011, the average subsidence rate is 124mm / a; In Panjin City, the area of (D) subsidence area is small, and the accumulative subsidence amount is -70 mm from 2007 to 2011, and the average subsidence rate is 23 mm / a. Based on the time series analysis of land subsidence, the settlement area of (A) in Dongguo Reed Plant has a tendency to increase, but the settlement rate has a tendency to slow down, while the settling rate of (B) in Huanxiling and (C) in Xibaqianxiang fluctuates wave-like and gradually slows down. The settlement area of (B) in Huanxiling gradually becomes smaller, and the subsidence area of (C) in Xibaqianxiang has an increasing trend, while the settling rate of (D) in Panjin fluctuates in waves. The twisted-cut tree method and the least cost flow method are used to study the phase diagram wrapping. The first difference method unwraps the interferometric phase directly and determines the scaling factor of the simulated terrain phase by the least square method so as to improve the terrain phase of the simulation. The second method does not need to unwrap the interference phase (not the differential phase), even if the interference phase can not be unwrapped successfully, the reasonable results can be calculated. The results obtained by two different unwrapping methods are basically consistent. In the four subsidence areas of (A), Huanxiling (B), Xibaqianxiang (C) and Panjin (D) in Dongguo Reed Factory, The maximum shape variables obtained by twisted-tangential tree method are -176mm-88mm-99mm and -28mm, respectively, and the maximum shape variables by the least cost flow method are -169mm / -78mm / -105mm and -16mm respectively. The results obtained by the least cost flow method are generally smaller than those obtained by the twisted-tangent tree method. The branch cutting tree unwrapping method is suitable for the calculation of unwrapping in plain area, and the minimum cost flow method is suitable for the global optimization, and is suitable for the unwrapping in mountainous and plain areas.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：P642.26
【参考文献】
相关期刊论文 前10条
1 吕乐婷;陈圣波;车大为;刘银萍;;D-InSAR原理及其数据处理流程[J];吉林大学学报(地球科学版);2008年S1期
2 李德仁,周月琴,马洪超;卫星雷达干涉测量原理与应用[J];测绘科学;2000年01期
3 荆燕,王建军,张景发,张红,蒋林根;D-InSAR技术在地震同震形变研究中的应用[J];测绘科学;2004年02期
4 李新武,郭华东,廖静娟,王长林,范典;基于快速傅立叶变换的干涉SAR基线估计[J];测绘学报;2003年01期
5 游新兆,乔学军,王琪,杜瑞林;合成孔径雷达干涉测量原理与应用[J];大地测量与地球动力学;2002年03期
6 路旭,匡绍君,贾有良,黄立人;用INSAR作地面沉降监测的试验研究[J];大地测量与地球动力学;2002年04期
7 王超,张红,于勇,刘智;雷达差分干涉测量[J];地理学与国土研究;2002年03期
8 姚国清;母景琴;;D-InSAR技术在地面沉降监测中的应用[J];地学前缘;2008年04期
9 单新建,马瑾,王长林,柳稼航;利用差分干涉雷达测量技术(D-InSAR)提取同震形变场[J];地震学报;2002年04期
10 徐华平,周荫清,李春升;星载干涉SAR中的基线问题[J];电子学报;2003年03期
相关博士学位论文 前1条
1 赵超英;差分干涉雷达技术用于不连续形变的监测研究[D];长安大学;2009年
本文编号：2214910
本文链接：https://www.wllwen.com/jingjilunwen/quyujingjilunwen/2214910.html
