| 摘要: |
| 在合成孔径雷达(SAR)海面风场反演中,基于风条纹影像纹理特征的海面风向反演方法精度高,但是依赖于图像风条纹的存在,而外部风向信息与SAR资料时空分辨率不易匹配、精度较低,从而影响大面积、高分辨率海面风场反演的精度。针对此问题,提出一种将SAR图像风条纹线性纹理特征与外部风向信息相结合的星载SAR海面风向获取方法,在SAR影像线性纹理特征明显的区域采用二维连续小波变换得到高精度的海面风向,其余区域采用与之时空相匹配的数值预报模式风向填充；并利用地球物理模型函数进一步得到海面风速,进而实现高精度、大范围海面风场的反演。为验证本文方法的有效性,利用ENVISAT/ASAR数据进行风场反演试验,并将反演结果与浮标实测数据进行比对。结果表明：在线性纹理特征明显的区域,小波方法的反演精度优于FFT法和数值预报模式风向；外部风向精度略低,但与SAR观测资料时空匹配性较好,弥补了风条纹风向的不足,二者结合为星载SAR海面风场反演的业务化应用提供了支持。 |
| 关键词: 星载SAR 小波变换 外部风向 海面风场 |
| DOI： |
| 分类号:P732.7 |
| 基金项目:国家自然科学基金(批准号：41475019,41375029) |
|
| Ocean Wind Retrieval from Space-borne SAR Images based on Wind Streaks and External Wind Direction |
|
CHENG Yu-xin,AI Wei-hua,KONG Yi and ZHAO Xian-bin
|
|
Institute of Meteorology and Oceanography,PLA University of Science and Technology,Institute of Meteorology and Oceanography,PLA University of Science and Technology,Institute of Meteorology and Oceanography,PLA University of Science and Technology,Institute of Meteorology and Oceanography,PLA University of Science and Technology
|
| Abstract: |
| The mothod of ocean wind direction retrieval based on wind streaks texture feature have high precision , which depends on the wind streaks . And external wind direction of low precision is not easy to match with the SAR data at space and time resolution , thus affecting large area, high resolution sea surface wind field retrieval accuracy . A mothed of ocean wind retrieval from space-borne SAR images was proposed , which combines wind streaks texture feature with external wind direction . Using two-dimensional continuous wavelet transform to get high precision of the surface wind direction in the area of SAR image which texture feature obviouslyS, the rest of area was fill by wind direction from numerical weather prediction model . C-band empirical model is used to calculate wind speed , then realize sea surface wind field retrieval of high precision, large range . Carring out wind field retrieval experiment though ENVISAT/ASAR sounding data and compare the retrieval result with the buoy measured data to verify the validity of the method in this paper.The results show that the retrieval accuracy of the wavelet is better than the FFT method and the numerical prediction model , which is matching on the space and time with the SAR observation data and make up for theSinadequacy of the wind direction of wind streaks . A combination will provide business application support for ocean wind field retrieval from space-borne SAR images. |
| Key words: Space-borne SAR wavelet transform External Wind Direction Ocean wind field |