测绘学报 ›› 2015, Vol. 44 ›› Issue (6): 616-624.doi: 10.11947/j.AGCS.2015.20140374

• 大地测量学与导航 • 上一篇    下一篇

长江口EnviSat测高数据的波形分类重构分析

彭福凯1,2, 沈云中1,2   

  1. 1. 同济大学测绘与地理信息学院, 上海 200092;
    2. 同济大学空间信息科学及可持续发展应用中心, 上海 200092
  • 收稿日期:2014-07-09 修回日期:2014-11-25 出版日期:2015-06-20 发布日期:2015-07-28
  • 作者简介:彭福凯(1989—),男,硕士,研究方向为雷达卫星高度计数据处理和应用。E-mail: pengfukai263@163.com
  • 基金资助:
    国家973计划(2012CB957703); 国家自然科学基金(41274035; 41474017)

Analysis of EnviSat Altimetric Data around Yangtze Estuary by Waveform Retracking

PENG Fukai1,2, SHEN Yunzhong1,2   

  1. 1. College of Surveying and Geo-informatics, Tongji University, Shanghai 200092, China;
    2. Center for Spatial Information Science and Sustainable Development, Tongji University, Shanghai 200092, China
  • Received:2014-07-09 Revised:2014-11-25 Online:2015-06-20 Published:2015-07-28
  • Supported by:
    The National Basic Research Program of China(973 Program)(No.2012CB957703);The National Natural Science Foundation of China (Nos. 41274035;41474017)

摘要: 采用波形分类重构算法处理EnviSat卫星从2002年10月至2010年5月在长江口近岸海域28 °N~32 °N、121 °E~125 °E范围内的波形数据。该区域内海洋波形、波形后缘前端出现峰值的波形、波形后缘后端出现峰值的波形、似镜面波形和复杂波形分别占89.03%、2.95%、0.45%、3.31%和4.26%。根据不同的波形类别采用不同波形算法进行波形重构。同时,分析了不同重构算法之间的系统偏差,并据此确定OCOG算法、Threshold算法和子波形算法的最优阈值水平分别为65%、45%和50%。重构结果表明,波形分类重构算法优于其他波形重构算法,能有效改善原始海面高的精度,改善程度在16.62%~53.86%之间。此外,重构后交叉点差值小于重构前的交叉点差值,与轨迹P089、P411形成的交叉点的海面高差值由1 m降低到25 cm左右,其余交叉点的差值均在2~6 cm。

关键词: 卫星测高, 波形分类, 聚类分析, 波形重构

Abstract: The waveform data of EnviSat satellite (28°N~32°N, 121°E~125°E) from October 2002 to May 2010 is processed by waveform retracking methodology based on waveform classification. The waveforms in this region are classified into ocean, pre-peak, post-peak, quasi-specular and complex waveforms, with percentages of 89.03%, 2.95%, 0.45%, 3.31% and 4.26%, respectively. For each waveform, correspondent waveform retracking methodology is used. Meanwhile, it is founded that there exist system biases between the sea surface heights(SSH) calculated by different waveform retracking methodologies and calculate optimum threshold level of OCOG, threshold and sub-waveform retracker, with 65%, 45% and 50% respectively. The results of waveform retracking show that the waveform retracking methodology based on waveform classification are prior to any single retracker, which could improve the precision of SSH from 16.62% to 53.86%. In addition, the crossover discrepancies are smaller after using waveform retracking. Specifically, the crossover discrepancies of pass 089 and pass 411 are declined from 1 m to 25 cm or so and the other crossover discrepancies are all around 2~6 cm.

Key words: satellite altimetry, waveform classification, cluster analysis, waveform retracking

中图分类号: