Simulation analysis for remote sensing inversion of wavelength and water depth by the Fast Fourier Transform method
-
摘要: 利用海浪波峰和波谷位置的遥感影像信息差异,可以基于快速傅里叶变换(FFT)方法反演波长,进而反演近岸水深.本文采用理想波面数据和数值模拟波面数据代替遥感资料进行仿真研究,讨论资料分辨率和子图长度对海浪波长及水深反演的影响.研究结果表明:低分辨率资料反演波长和水深的效果差,但是当资料分辨率达到一定要求时,再提高资料分辨率对波长和水深反演结果没有影响.当波长不存在空间变化时,子图越大,波长反演误差越小.在非均匀变化的地形上,子图长度为4~8倍波长、并且小于或等于地形变化尺度时,波长和水深反演误差小,子图太大或太小时,波长和水深反演误差都增大.Abstract: Using the difference in remote sensing reflectivity between wave crest and trough, wavelengths can be inversed by the Fast Fourier Transform (FFT) method to derive the nearshore water depth. Remote sensing images with higher resolution are generally thought to induce less error in marine information inversion. In this study, remote sensing images were replaced by elevation data from ideal equations and numerical simulations to study the effect of data resolution on wavelength inversions and water depths. The results show that low-resolution data result in significant errors in the wavelengths and water depths. But variations in resolution make no difference in reversing the wavelengths and water depths as they reach certain levels. The sub-image size in the FFT method was also studied by simulation analysis. Larger sub-images generate less error in wavelength inversion if the wavelength doesn't vary spatially. On uneven topography, errors from wavelength and water depth inversions are small if the sub-image size is 4-8 times the wavelength and not more than the topography variation; however, when the sub-image is too large or too small, the errors will increase.
-
表 1 波长为70m的理想波面资料反演的波长
Tab. 1 Wavelengths from inversions by elevation data at a wavelength of 70 m
m 子图长度 不同分辨率资料反演的波长 0.25m 0.5m 1m 2m 4m 8m 16m 32m 64m 64m 64 64 64 64 64 64 64 64 - 128m 64 64 64 64 64 64 64 64 - 256m 64 64 64 64 64 64 64 64 - 512m 73 73 73 73 73 73 73 73 512 1 024m 68 68 68 68 68 68 68 68 1 024 表 2 波长为90m的理想波面资料反演的波长
Tab. 2 Wavelengths from inversions by elevation data at a wavelength of 90 m
m 子图长度 不同分辨率资料反演的波长 0.25m 0.5m 1m 2m 4m 8m 16m 32m 64m 64m 64 64 64 64 64 64 64 64 - 128m 128 128 128 128 128 128 128 128 128 256m 85 85 85 85 85 85 85 85 256 512m 85 85 85 85 85 85 85 85 256 1 024m 93 93 93 93 93 93 93 93 205 表 3 波长为110m的理想波面资料反演的波长
Tab. 3 Wavelengths from inversions by elevation data at a wavelength of 110 m
m 子图长度 不同分辨率资料反演的波长 0.25m 0.5m 1m 2m 4m 8m 16m 32m 64m 64m - - - - - - - - - 128m 128 128 128 128 128 128 128 128 - 256m 128 128 128 128 128 128 128 128 128 512m 102 102 102 102 102 102 102 102 128 1 024m 114 114 114 114 114 114 114 114 146 表 4 组合波动(振幅之比1.2)的理想波面资料反演的波长
Tab. 4 Wavelengths from inversions by elevation data including two fluctuations with an amplitude ratio of 1.2
m 子图长度 不同分辨率资料反演的波长 0.25m 0.5m 1m 2m 4m 8m 16m 32m 64m 64m 64 64 64 64 64 64 64 64 - 128m 64 64 64 64 64 64 64 64 128 256m 64 64 64 64 64 64 64 128 128 512m 102 102 102 102 102 102 102 102 170 1 024m 68 68 68 68 68 68 68 102 170 表 5 组合波动(振幅之比1.6)的理想波面资料反演的波长
Tab. 5 Wavelengths from inversions by elevation data including two fluctuations with an amplitude ratio of 1.6
m 子图长度 不同分辨率资料反演的波长 0.25m 0.5m 1m 2m 4m 8m 16m 32m 64m 64m 64 64 64 64 64 64 64 64 - 128m 64 64 64 64 64 64 64 64 - 256m 64 64 64 64 64 64 64 64 - 512m 73 73 73 73 73 73 73 73 512 1 024m 68 68 68 68 68 68 68 68 1 024 表 6 在斜坡地形上利用数值模拟仿真数据反演水深的误差统计
Tab. 6 Errors of water depths from inversions using the simulated elevation data on a slope
子图长度 0-3 m水深段的误差统计 3-12 m水深段的误差统计 12-20 m水深段的误差统计 平均绝对
误差/m平均相对
误差/%平均绝对
误差/m平均相对
误差/%平均绝对
误差/m平均相对
误差/%128m 0.45 19.59 1.71 23.19 7.50 48.13 256m 0.29 12.34 1.65 19.02 2.24 15.22 512m 0.23 10.34 0.56 7.62 1.77 11.44 1 024m 0.48 21.43 0.35 4.89 1.21 7.20 表 7 在沙坝地形上利用数值模拟仿真数据反演水深的误差统计
Tab. 7 Errors of water depths from inversions using the simulated elevation data on bars
沙坝尺寸 子图长度128 m时 子图长度256 m时 子图长度512 m时 子图长度1 024 m时 平均绝对
误差/m平均相对
误差/%平均绝对
误差/m平均相对
误差/%平均绝对
误差/m平均相对
误差/%平均绝对
误差/m平均相对
误差/%125m 1.37 27.91 0.79 16.37 1.27 30.12 3.05 72.04 250m 1.37 28.87 0.73 15.82 1.08 20.93 2.99 66.83 500m 0.95 19.39 0.70 15.36 0.47 10.64 1.21 23.14 1 000m 1.47 21.16 1.26 16.27 0.62 10.29 1.09 19.43 -
[1] 申家双, 潘时祥.沿岸水深测量技术方法的探讨[J].海洋测绘, 2002, 22(6):60-65. doi: 10.3969/j.issn.1671-3044.2002.06.016 [2] 周高伟, 李英成, 任延旭, 等.低空无人机双介质水下礁盘深度测量试验与分析[J].测绘学报, 2015, 44(5):548-554. http://d.old.wanfangdata.com.cn/Periodical/chxb201505012 [3] 刘焱雄, 郭锴, 何秀凤, 等.机载激光测深技术及其研究进展[J].武汉大学学报(信息科学版), 2017, 42(9):1185-1194. http://d.old.wanfangdata.com.cn/Periodical/whchkjdxxb201709001 [4] 张鹰, 张东, 王艳姣, 等.含沙水体水深遥感方法的研究[J].海洋学报, 2008, 30(1):51-58. doi: 10.3321/j.issn:0253-4193.2008.01.007 [5] 沈婕, 苏昆, 张鹰, 等.基于遥感反演水深数据的测图技术研究[J].测绘科学, 2009, 34(4):180-181. http://d.old.wanfangdata.com.cn/Periodical/chkx200904065 [6] 张晓冬, 张文静, 朱首贤, 等.海口湾可见光遥感测深方法研究[J].海洋通报, 2016, 35(1):54-63. http://d.old.wanfangdata.com.cn/Periodical/hytb201601008 [7] ALPERS W, HENNINGS I. A theory of the imaging mechanism of underwater bottom topography by real and synthetic aperture radar[J]. Journal of Geophysical Research Oceans, 1984, 89(C6):10529-10546. doi: 10.1029/JC089iC06p10529 [8] 黄韦艮, 傅斌, 周长宝, 等.星载SAR遥感浅海水下地形的最佳海况模拟仿真[J].自然科学进展:国家重点实验室通讯, 2000, 10(7):642-649. http://d.old.wanfangdata.com.cn/Periodical/zrkxjz200007008 [9] 范开国, 黄韦艮, 贺明霞, 等. SAR浅海水下地形遥感研究进展[J].遥感技术与应用, 2008, 23(4):479-485. http://d.old.wanfangdata.com.cn/Periodical/ygjsyyy200804022 [10] 滕惠忠, 熊显名, 李海滨, 等.遥感水深反演海图修测应用研究[J].海洋测绘, 2009, 29(6):21-25. doi: 10.3969/j.issn.1671-3044.2009.06.006 [11] 叶安乐, 李凤岐.物理海洋学[M].山东青岛:青岛海洋大学出版社, 1992. [12] PLESKACHEVSKY A, LEHNER S, HEEGE T, et al. Synergy and fusion of optical and synthetic aperture radar satellite data for underwater topography estimation in coastal areas[J]. Ocean Dynamics, 2011, 61(12):2099-2120. doi: 10.1007/s10236-011-0460-1 [13] DANILO C, MELGANI F. Wave period and coastal bathymetry using wave propagation on optical images[J]. IEEE Transactions on Geoscience & Remote Sensing, 2016, 54(11):6307-6319. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=2d5ec58af49fa325408450e2b9ffaba8 [14] BELL P. Shallow water bathymetry derived from an analysis of x-band marine radar images of waves[J]. Coastal Eng, 1999, 37:513-527. doi: 10.1016/S0378-3839(99)00041-1 [15] LYU L, CHANG H. Remotely sensing in detecting the water depths and bed load of shallow waters and their changes[J]. Ocean Engineering, 2005, 32(10):1174-1198. doi: 10.1016/j.oceaneng.2004.12.005 [16] 陈台颖.频谱分析应用与决定卫卫星影像中的波向及水深推估[D].台湾新竹: 台湾交通大学, 2013. [17] 柯绅彦.应用卫卫星影像决定波向线及估算水深之初步探讨[D].台湾新竹: 台湾交通大学, 2012. [18] LI J, ZHANG H, HOU P, et al. Mapping the bathymetry of shallow coastal water using single-frame fineresolution optical remote sensing imagery[J].海洋学报(英文版), 2016, 35(1):60-66. doi: 10.1007/s13131-016-0797-x [19] POUPARDIN A, IDIER D, MICHELE M D, et al. Water Depth Inversion From a Single SPOT-5 Dataset[J]. IEEE Transactions on Geoscience & Remote Sensing, 2016, 54(4):2329-2342. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=2802e67b990ca0eef1b477a19d2107e3 [20] 傅斌, 黄韦艮, 周长宝, 等.星载SAR浅海水下地形和水深测量模拟仿真——水下地形高度、坡度和方向与可测水深分析[J].海洋学报, 2001, 23(1):35-42. doi: 10.3321/j.issn:0253-4193.2001.01.005 [21] KUO Y Y, LYU L G, KAO I L. Directional spectrum analysis and statistics obtained from ERS-1 SAR wave images[J]. Ocean Engineering, 1999, 26(11):1125-1144. doi: 10.1016/S0029-8018(98)00058-4 [22] KIRBY J T, WEI G, CHEN Q, et al. Funwave 1.0: Fully nonlinear boussinesq wave model-documentation and user's manual[R]. DE: University of Delaware, 1998. [23] SHI F Y, KIRBY J T, TEHRANIRAD B, et al. FUNWAVE-TVD, documentation and users' manual (CACR-11-03)[R]. DE: University of Delaware, 2011. [24] BERKHOFF J C W, BOOY N, RADDER A C. Verification of numerical wave propagation models for simple harmonic linear water waves[J]. Coastal Engineering, 1982, 6(3):255-279. doi: 10.1016/0378-3839(82)90022-9 [25] 王颖.黄海陆架辐射沙脊群[M].北京:中国环境科学出版社, 2002. [26] 陈玮彤, 张东, 施顺杰, 等.江苏中部淤泥质海岸岸线变化遥感监测研究[J].海洋学报, 2017, 39(5):138-148. doi: 10.3969/j.issn.0253-4193.2017.05.013