-
摘要:
海面高度(SSH)在海洋学和气象学等领域中具有重要意义。针对目前缺乏利用中国自主研发风云三号E星(FY-3E)的全球导航卫星掩星探测仪-Ⅱ型(GNOS-Ⅱ)特有的不均匀分布时延-多普勒(DDM)数据实现海面测高研究的问题，采用DDM海面高度反演技术，使用丹麦DTU18模型和海潮模型验证反演精度，综合传统物理模型和机器学习模型，分别对星载北斗导航卫星系统(BDS)和全球定位系统(GPS)反射信号的海面高度反演性能进行评估。结果显示，由于BDS B1Ⅰ信号的码片分辨率高于GPS，使用BDS的物理模型进行全球海面高度反演的最大平均绝对误差(MAE)约为3.0 m，明显优于GPS反演结果(最大MAE约为5.0 m)。通过随机森林(RF)和卷积神经网络(CNN)模型，GPS和BDS均能实现较好的反演结果，最佳MAE均约为0.4 m。与英国TDS-1数据的反演结果相比，GPS反演精度在物理模型上提高约15%，验证了FY-3E GNOS-Ⅱ的全球导航卫星系统反射信号(GNSS-R)遥感数据的有效性。研究成果对于推广国产FY-3E的GNSS-R海面测高应用具有重要意义。
-
关键词:
- 全球导航卫星系统反射信号 /
- 风云三号E星 /
- 海面高度反演 /
- 机器学习 /
- 特征选择
Abstract:Sea surface height (SSH) is of great significance in fields such as oceanography and meteorology. Given the current absence of research utilizing the unevenly distributed delay-Doppler map (DDM) data from China’s independently developed FengYun-3E (FY-3E) global navigation satellite system occultation sounderⅡ (GNOS-Ⅱ), this paper employs DDM-based SSH retrieval technology and validates the retrieval accuracy using the Danish DTU18 global mean SSH model and tidal model. The SSH retrieval performance of reflected signals from the global positioning system (GPS) and the spaceborne Beidou navigation satellite system (BDS) is evaluated individually in this research using both conventional physical models and machine learning techniques. The results demonstrate that due to the higher chip resolution of BDS B1Ⅰ signals compared to GPS, the maximum mean absolute error (MAE) of the global SSH inversion using the physical model of BDS is about 3.0 m, which is significantly better than that of the GPS (maximum MAE is about 5.0 m). Both GPS and BDS may produce good inversion results using the random forest (RF) and convolutional neural network (CNN) models; their best MAE is approximately 0.4 m. Compared with the inversion results of the UK TDS-1 data, the GPS inversion accuracy is improved by about 15% in the physical model, which verifies the validity of the global navigation satellite system-reflectometry (GNSS-R) remote sensing data of the FY-3E GNOS-Ⅱ. The research findings in this paper are of great significance for promoting the application of domestic FY-3E GNSS-R SSH measurements.
-
图 1 FY-3E数据在2022年8月25日镜面反射点分布
Figure 1. Distribution of specular points for FY-3E data on August 25th, 2022
图 2 FY-3E反射产品中典型的海面DDM图像
Figure 2. Typical sea surface DDM images in FY-3E reflection products
图 3 星载GNSS-R测高几何路径延迟模型
Figure 3. Geometric path delay model for spaceborne GNSS-R altimetry
图 4 星载GNSS-R测高的几何原理
Figure 4. Geometric schematic diagram of spaceborne GNSS-R altimetry
图 5 插值后的多普勒频率为0的时延一维功率谱切片
Figure 5. One-dimensional power spectrum slice with zero Doppler shift after interpolation
图 7 机器学习数据处理流程及RF训练模型结构
Figure 7. Machine learning data processing process and random forest training model structure
图 8 单个文件海面高度反演结果（BDS）
Figure 8. Single file sea surface height inversion results (BDS)
图 9 单个文件海面高度反演结果（GPS）
Figure 9. Single file sea surface height inversion results (GPS)
图 10 发射机为 BDS 时的误差(残差)地理分布（物理模型，2022年8月7日、12日和17日）
Figure 10. Geographical distribution of errors (residual) when transmitter is BDS (physical model, 2022-08-07, 2022-08-12, 2022-08-17)
图 11 发射机为 GPS 时的误差(残差)地理分布（物理模型，2022年8月7日、12日和17日）
Figure 11. Geographical distribution of errors (residual) when transmitter is GPS (physical model, 2022-08-07, 2022-08-12, 2022-08-17)
图 12 发射机为 BDS 时的误差(残差)地理分布（RF模型，2022年8月7日、12日和17日）
Figure 12. Geographical distribution of errors (residual) when transmitter is BDS (RF model, 2022-08-07,2022-08-12, 2022-08-17)
图 13 发射机为 GPS 时的误差(残差)地理分布（RF模型，2022年8月7日、12日和17日）
Figure 13. Geographical distribution of errors (residual) when transmitter is GPS (RF model, 2022-08-07,2022-08-12, 2022-08-17)
图 15 发射机为 BDS 时的误差(残差)地理分布（CNN模型，2022年8月7日、12日和17日）
Figure 15. Geographical distribution of errors (residual) when transmitter is BDS (CNN model, 2022-08-07,2022-08-12, 2022-08-17)
图 16 发射机为 GPS 时的误差(残差)地理分布（CNN模型，2022年8月7日、12日和17日）
Figure 16. Geographical distribution of errors (residual) when transmitter is GPS (CNN model, 2022-08-07, 2022-08-12, 2022-08-17)
表 1 BDS B1Ⅰ信号与GPS L1信号的比较
Table 1. Comparison between BDS B1Ⅰ signals and GPS L1 signals
信号 频点/MHz 带宽/MHz 波长/m 调制方式 伪码速率/MHz 码片个数 码片宽度/μs BDS B1Ⅰ 1561.098 4.092 0.192 BPSK 2.046 2046 0.489 GPS L1 1575.42 2.046 0.191 BPSK 1.023 1023 0.978 
下载: 导出CSV
表 2 数据质量控制后的数据数量
Table 2. Amount of data after data quality control
模型 BDS/
GPS训练集数据量 反演数据量 2022-08-01—
2022-08-062022-08-07 2022-08-12 2022-08-17 机器学
习模型BDS 266954 60912 50339 57926 GPS 260100 54432 52695 45151 物理模型 BDS 31784 24447 28800 GPS 34135 31823 30368
下载: 导出CSV
表 3 物理模型反演结果精度
Table 3. Accuracy of physical model inversion results
反演日期 MAE/m RMSE/m BDS GPS BDS GPS 2022-08-07 2.635 5.539 3.702 6.801 2022-08-12 2.812 5.260 3.939 6.468 2022-08-17 3.501 4.702 4.631 5.867
下载: 导出CSV
表 4 RF模型反演结果的精度
Table 4. Accuracy of RF model inversion results
预测日期 MAE/m RMSE/m BDS GPS BDS GPS 2022-08-07 0.439 0.471 0.663 0.816 2022-08-12 0.531 0.597 0.743 0.989 2022-08-17 0.481 0.467 0.813 0.908
下载: 导出CSV
表 5 CNN模型反演结果精度
Table 5. Accuracy of CNN model prediction results
预测日期 MAE/m RMSE/m BDS GPS BDS GPS 2022-08-07 0.429 0.495 0.573 0.652 2022-08-12 0.578 0.593 0.745 0.766 2022-08-17 0.496 0.435 0.675 0.584
下载: 导出CSV
-
[1] 陈锐志, 王磊, 李德仁, 等. 导航与遥感技术融合综述[J]. 测绘学报, 2019, 48(12): 1507-1522.CHEN R Z, WANG L, LI D R, et al. A survey on the fusion of the navigation and the remote sensing techniques[J]. Acta Geodaetica et Cartographica Sinica, 2019, 48(12): 1507-1522(in Chinese). [2] MARTIN-NEIRA M. A passive reflectometry and interferometry system (PARIS): application to ocean altimetry[J]. ESA Journal, 1993, 17(4): 331-335. [3] 张云, 张杨阳, 孟婉婷, 等. 机载GNSS反射信号海面测高模型的研究[J]. 海洋学报, 2020, 42(3): 149-156. doi: 10.3969/j.issn.0253-4193.2020.03.014ZHANG Y, ZHANG Y Y, MENG W T, et al. Research on sea surface altimetry model of airborne GNSS reflected signal[J]. Haiyang Xuebao, 2020, 42(3): 149-156(in Chinese). doi: 10.3969/j.issn.0253-4193.2020.03.014 [4] ZHANG Y, TIAN L M, MENG W T, et al. Feasibility of code-level altimetry using coastal BeiDou reflection (BeiDou-R) setups[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2015, 8(8): 4130-4140. doi: 10.1109/JSTARS.2015.2446684 [5] ZHANG Y, LI B B, TIAN L M, et al. Phase altimetry using reflected signals from BeiDou GEO satellites[J]. IEEE Geoscience and Remote Sensing Letters, 2016, 13(10): 1410-1414. doi: 10.1109/LGRS.2016.2578361 [6] 李黄, 夏青, 尹聪, 等. 我国GNSS-R遥感技术的研究现状与未来发展趋势[J]. 雷达学报, 2013, 2(4): 389-399. doi: 10.3724/SP.J.1300.2013.13080LI H, XIA Q, YIN C, et al. The current status of research on GNSS-R remote sensing technology in China and future development[J]. Journal of Radars, 2013, 2(4): 389-399(in Chinese). doi: 10.3724/SP.J.1300.2013.13080 [7] 刘风玲, 张云, 孟婉婷, 等. UK-DMC卫星接收机GNSS反射信号的应用分析[J]. 遥感信息, 2015, 30(1): 90-95. doi: 10.3969/j.issn.1000-3177.2015.01.015LIU F L, ZHANG Y, MENG W T, et al. Application analysis of airborne GNSS reflected data from UK-DMC satellite receiver[J]. Remote Sensing Information, 2015, 30(1): 90-95(in Chinese). doi: 10.3969/j.issn.1000-3177.2015.01.015 [8] CLARIZIA M P, RUF C, CIPOLLINI P, et al. First spaceborne observation of sea surface height using GPS-reflectometry[J]. Geophysical Research Letters, 2016, 43(2): 767-774. doi: 10.1002/2015GL066624 [9] MASHBURN J, AXELRAD P, LOWE S T, et al. Global ocean altimetry with GNSS reflections from TechDemoSat-1[J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 56(7): 4088-4097. doi: 10.1109/TGRS.2018.2823316 [10] 张云, 马德皓, 孟婉婷, 等. 基于TechDemoSat-1卫星的GPS反射信号海面高度反演[J]. 北京麻豆精品秘 国产传媒学报, 2021, 47(10): 1941-1948.ZHANG Y, MA D H, MENG W T, et al. Sea surface height inversion of GPS reflected signal based on TechDemoSat-1 satellite[J]. Journal of Beijing University of Aeronautics and Astronautics, 2021, 47(10): 1941-1948(in Chinese). [11] WANG Q, ZHENG W, WU F, et al. A new GNSS-R altimetry algorithm based on machine learning fusion model and feature optimization to improve the precision of sea surface height retrieval[J]. Frontiers in Earth Science, 2021, 9: 730565. doi: 10.3389/feart.2021.730565 [12] ZHANG Y, HUANG S, HAN Y L, et al. Machine learning methods for spaceborne GNSS-R sea surface height measurement from TDS-1[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 15: 1079-1088. [13] ZHANG Y, LU Q, JIN Q, et al. Global sea surface height measurement from CYGNSS based on machine learning[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 16: 841-852. [14] ZHANG P, HU X Q, LU Q F, et al. FY-3E: the first operational meteorological satellite mission in an early morning orbit[J]. Advances in Atmospheric Sciences, 2022, 39(1): 1-8. doi: 10.1007/s00376-021-1304-7 [15] SUN Y Q, WANG X Y, DU Q F, et al. The status and progress of Fengyun-3E GNOS Ⅱ mission for GNSS remote sensing[C]//Proceedings of the IEEE International Geoscience and Remote Sensing Symposium. Piscataway: IEEE Press, 2019: 5181-5184. [16] YANG G L, BAI W H, WANG J S, et al. FY3E GNOS Ⅱ GNSS reflectometry: mission review and first results[J]. Remote Sensing, 2022, 14(4): 988. doi: 10.3390/rs14040988 [17] ANDERSEN O B, KNUDSEN P. The DTU2010MSS mean sea surface in the arctic-for and with Cyrosat-2 data[J]. Cryosat Validation Workshop, 2011, 693: 28. [18] PICCIONI G, ANDERSEN O B, STENSENG L. SAR altimetry for mean sea surface determination in the arctic DTU15MSS[C]//Proceedings of the Sentinel-3 for Science Workshop. Venice : European Space Agency, 2015: 75. [19] GARRISON J L, KATZBERG S J. The application of reflected GPS signals to ocean remote sensing[J]. Remote Sensing of Environment, 2000, 73(2): 175-187. doi: 10.1016/S0034-4257(00)00092-4 -
下载:
点击查看大图
计量
- 文章访问数: 824
- HTML全文浏览量: 92
- PDF下载量: 5
- 被引次数: 0
