重轨卫星InSAR技术在多云雨山区DEM生成及更新中的应用研究
发布时间：2018-11-12 17:56
【摘要】：数字高程模型(Digital Elevation Model,DEM)作为一种重要的地理空间数据,在国民经济和国防建设以及人文和自然科学领域有着广泛应用。就我国而言,目前还没有覆盖全球的高精度DEM,加快推进我国海洋国土乃至全球的高精度DEM获取是《全国基础测绘中长期规划纲要(2015——2030年)》的明确要求。野外数字测量、地形图扫描数字化、摄影测量和机载Lidar等方法也可获取局部高分辨率DEM,但均耗时耗力,易受云雾雨雪天气影响,不适于大区域尤其是境外DEM的获取。地面GPS的采样点非常稀疏,成本很高,限制了其发展。重轨卫星InSAR技术具有主动成像、穿云透雾、全天时工作等特点,成为大范围获取DEM的研究热点。并且我国高分三号卫星于2017年正式投入使用,获取了首批高分辨率SAR影像,可为通过InSAR技术获取国内乃至全球的高精度DEM提供数据支持。为获取更高精度的InSAR DEM,国内外学者开展了诸多尝试,相关研究集中在三个方面:一是外部DEM(如SRTM DEM)辅助地形复杂区域干涉相位进行相位解缠,降低解缠难度,减小解缠误差,但没考虑大气相位噪声的影响,也没对完全失相干的水域进行高精度探测及高程赋值;二是开展多基线InSAR联合估计DEM,通过冗余数据增加观测次数以减小观测误差,提高DEM的精度,但没充分考虑消除大气相位噪声的影响;三是时间序列InSAR技术正趋成熟,而国内外学者多集中于提高反演地表形变的精度,利用反演的高程残差获取城区建筑物的高度信息也有少量研究,但基于高程残差更新外部低精度DEM以获取更高精度DEM的研究目前尚未见到。针对以上问题,本文在常年云雨天气、地形复杂、植被茂密的云南省昆明市高山地开展重轨卫星单基线InSAR、多基线InSAR获取DEM研究;在地势复杂、雨雪较多的辽宁省葫芦岛市北部低山丘陵区开展时间序列InSAR更新DEM研究;用国家1:50000 DEM一级数据评定实验结果的高程精度。主要工作如下:一、开展基于外部DEM辅助解缠的单基线InSAR技术研究。提出利用外部DEM恢复干涉相位整周模糊数的方法;改进“二轨”差分辅助相位解缠的方法,称之为“差分补偿”法:一是基于差分相位的条纹频率初次精化基线,二是对影像进行水域探测及高程赋值。选用2景昆明市高山地区ALOS-2 PALSAR-2影像(20160701、20160729)开展常规InSAR、外部DEM恢复干涉相位整周模糊数InSAR和“差分补偿”InSAR的对比实验研究。结果表明:三种方法获取的DEM精度均满足国家1:100000DEM三级标准;外部DEM恢复整周模糊数InSAR DEM精度最高,差分补偿InSAR DEM次之,常规InSAR DEM最低;水域探测结果较为理想,精度较高。二、开展多基线InSAR相位累积方法研究。提出干涉相位累积法和差分相位累积补偿法;提出时序相位定权叠加法。选用4景昆明市高山地区ALOS-2 PALSAR-2影像(20160701、20160715、20160729、20160909)开展实验研究。结果表明:相位累积法可显著减弱大气相位噪声;由4景影像组构的六基线相位累积获取的InSAR DEM精度基本满足美国DTED-1标准和国家1:50000 DEM三级标准。三、开展时间序列InSAR更新DEM方法研究。时间序列InSAR技术通过提取高密度高质量的稳定点目标,有效地克服了传统InSAR的大气效应、时间去相干、基线去相干等因素的影响,可高精度地获取点目标上的地表形变和外部DEM高程残差信息。对点目标的高程残差做空间插值处理,得到全监测区的高程残差信息,以此更新外部DEM高程。选用22景葫芦岛市低山丘陵区ALOS-1 PALSAR-1影像开展试验性探究,提取了图像像点占比约为6.34%的高相干点目标及其高程残差,并进行克里金插值。选取范围为9 km×9 km、地势陡峭、植被较密、无人工建/构筑物、无地面沉降的小区域进行重点分析。结果表明:SRTM DEM更新后,高程精度提高了15.6%。
[Abstract]:Digital Elevation Model (DEM), as an important geospatial data, is widely used in national economy and national defense construction, as well as in the fields of human and natural science. As far as our country is concerned, the global high-precision DEM has not yet been covered, and the high-precision DEM acquisition of China's marine land and even the whole world is a clear requirement for the national basic and long-term plan for long-and long-term planning (2015 _ 2030). The local high-resolution DEM can also be obtained by field digital measurement, topographic map scanning digitization, photogrammetry, and on-board Lidar. However, it is time-consuming and easy to be affected by the cloud and snow weather, and is not suitable for the acquisition of large area, especially the overseas DEM. The ground GPS sampling point is very sparse, the cost is high, and its development is limited. The heavy-orbit satellite InSAR technology has the characteristics of active imaging, cloud-penetrating, full-day operation and so on, and becomes a hot spot for the large-scale acquisition of DEM. and the first batch of high-resolution SAR images is obtained by the high-resolution satellite 3 in 2017, and can be used for providing data support for high-precision DEM obtained at home and even in the world through the InSAR technology. In order to obtain the more accurate InSAR DEM, the domestic and foreign scholars have made a number of attempts, and the relevant research has focused on three aspects: one is the external DEM (such as the SRTM DEM) to assist in the phase unwrapping of the complex area interference phase, to reduce the difficulty of the solution and to reduce the unwrapping error. but the influence of the atmospheric phase noise is not taken into consideration, and the high-precision detection and the elevation assignment are not carried out on the fully distorted water area; secondly, the multi-baseline InSAR combined estimation DEM is carried out, the observation times are increased by the redundant data to reduce the observation error, and the accuracy of the DEM is improved, but the effect of eliminating the atmospheric phase noise is not fully considered; the three is that the time series InSAR technology is becoming more mature, and the domestic and foreign scholars are more focused on improving the accuracy of the inversion of the surface deformation, However, the method of updating the external low-precision DEM based on the elevation residual is not seen at present. In view of the above problems, in this paper, a single-baseline InSAR, a multi-baseline InSAR, a multi-baseline InSAR, is used to study the single-baseline InSAR and multi-baseline InSAR in the high mountainous area of Kunming, which is a complex and dense vegetation. The time series InSAR is used to update the DEM for the low mountain area in the northern part of Huludao City, Liaoning Province, and the elevation accuracy of the experimental results is evaluated by the first-level data of the national 1: 50000 DEM. The main work is as follows: 1. Carry out the single-baseline InSAR technology research based on the auxiliary unwrapping of the external DEM. This paper presents a method to recover interference phase integer ambiguity by using external DEM, and to improve the 鈥渢wo-track鈥，
本文编号：2327797
[Abstract]:Digital Elevation Model (DEM), as an important geospatial data, is widely used in national economy and national defense construction, as well as in the fields of human and natural science. As far as our country is concerned, the global high-precision DEM has not yet been covered, and the high-precision DEM acquisition of China's marine land and even the whole world is a clear requirement for the national basic and long-term plan for long-and long-term planning (2015 _ 2030). The local high-resolution DEM can also be obtained by field digital measurement, topographic map scanning digitization, photogrammetry, and on-board Lidar. However, it is time-consuming and easy to be affected by the cloud and snow weather, and is not suitable for the acquisition of large area, especially the overseas DEM. The ground GPS sampling point is very sparse, the cost is high, and its development is limited. The heavy-orbit satellite InSAR technology has the characteristics of active imaging, cloud-penetrating, full-day operation and so on, and becomes a hot spot for the large-scale acquisition of DEM. and the first batch of high-resolution SAR images is obtained by the high-resolution satellite 3 in 2017, and can be used for providing data support for high-precision DEM obtained at home and even in the world through the InSAR technology. In order to obtain the more accurate InSAR DEM, the domestic and foreign scholars have made a number of attempts, and the relevant research has focused on three aspects: one is the external DEM (such as the SRTM DEM) to assist in the phase unwrapping of the complex area interference phase, to reduce the difficulty of the solution and to reduce the unwrapping error. but the influence of the atmospheric phase noise is not taken into consideration, and the high-precision detection and the elevation assignment are not carried out on the fully distorted water area; secondly, the multi-baseline InSAR combined estimation DEM is carried out, the observation times are increased by the redundant data to reduce the observation error, and the accuracy of the DEM is improved, but the effect of eliminating the atmospheric phase noise is not fully considered; the three is that the time series InSAR technology is becoming more mature, and the domestic and foreign scholars are more focused on improving the accuracy of the inversion of the surface deformation, However, the method of updating the external low-precision DEM based on the elevation residual is not seen at present. In view of the above problems, in this paper, a single-baseline InSAR, a multi-baseline InSAR, a multi-baseline InSAR, is used to study the single-baseline InSAR and multi-baseline InSAR in the high mountainous area of Kunming, which is a complex and dense vegetation. The time series InSAR is used to update the DEM for the low mountain area in the northern part of Huludao City, Liaoning Province, and the elevation accuracy of the experimental results is evaluated by the first-level data of the national 1: 50000 DEM. The main work is as follows: 1. Carry out the single-baseline InSAR technology research based on the auxiliary unwrapping of the external DEM. This paper presents a method to recover interference phase integer ambiguity by using external DEM, and to improve the 鈥渢wo-track鈥，
本文编号：2327797
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2327797.html
