Journal of Geoinformation Science >
Wind Field Retrieval of South China Sea Based on GaussianFFT Method
Received date: 20151207
Request revised date: 20160505
Online published: 20161120
Copyright
Wind speed is a basic parameter of oceanography. It plays an important role in the interaction between ocean and atmosphere. Therefore, it is significant and necessary to obtain the wind datasets over the sea surface. However, due to the large area and complex condition, it is usually difficult to get the wind field data of South China to satisfy different demands in time. Conventional approaches, such as placing observation stations or buoys, are not only expensive but also dependent on the weather condition. Therefore it is urgently necessary to find other ways to get the wind datasets timely. ENSISAT ASAR, an allweather and alltime microwave radar sensor, could collect the realtime and dynamic information over sea surface, which provides a new approach for researchers to acquire wind field datasets over sea surface, especially for the waters with complicated conditions, such as South China Sea. In this paper, the GaussianFFT method is firstly applied to retrieve the wind field of South China Sea based on ASAR image. At first, the FFT spectrum of ASAR image is acquired with the FFT algorithm. Secondly, a “cigarshaped” twodimensional (2D) Gaussian function is fitted to the FFT spectrum to find the direction of wind streaks and further to obtain the wind direction which is perpendicular to it. In this experiment, the wind direction acquired from the ASAR image by the GaussianFFT algorithm also has a 180 ambiguity in direction. To resolve the 180 ambiguity, CCMP wind field datasets are taken into consideration to act as the wind field references. Besides, the wind direction computed with the GaussianFFT method is compared with the wind direction obtained by the PeakFFT method. Then, the optimal wind direction (GaussianFFT wind direction) is input into the CMOD4 and CMOD5 models to compute the wind speed values respectively. Through comparing the wind field retrieval results with the CCMP datasets, we proved that it is valid to retrieve wind direction from ASAR image with GaussianFFT algorithm and it is achievable to obtain wind speed value over South China Sea with CMOD4 model. The approach used to obtain the wind field in this paper is of great significance to provide guidance to the wind field inversion in other waters of South China Sea, especially in areas that are lack of field observations. In addition, it is also critical for other researches whose specialties are related to oceanography, as this approach could offer vital wind parameters to these researches.
Key words： South China Sea wind retrieval; ENVISAT ASAR; GaussianFFT; CMOD4 model; CMOD5 model
TIAN Siyu , HUANG Xiaoxia , LI Hongga , WANG Hao , LI Xia , CHENG Peng . Wind Field Retrieval of South China Sea Based on GaussianFFT Method[J]. Journal of Geoinformation Science, 2016 , 18(11) : 1544 1550 . DOI: 10.3724/SP.J.1047.2016.01544
Fig.1 Preprocessed SAR image图1 预处理后的SAR 影像 
Fig.2 CCMP wind data at UTC 12:00 and UTC 18:00 on April 16, 2006图2 CCMP 风场2006年4月16日12:00和18:00时（UTC）数据 
Fig.3 Flowchart of wind direction retrieval图3 风向反演流程图 
Tab.1 ASAR subimages’ wind direction retrieval results with respect to GaussianFFT method and PEAKFFT method (°)表1 ASAR子图像的高斯曲线拟合FFT方法和峰值FFT方法反演风向值(°) 
子图名  高斯曲线拟合FFT风向  峰值FFT风向  CCMP风向  高斯曲线拟合FFT风向与CCMP风向差值  峰值FFT风向与CCMP风向差值  

1  218.32  213.69  227.28  8.96  13.59  
2  222.01  225.00  226.19  4.18  1.19  
3  213.00  209.75  224.99  11.99  15.24  
4  222.82  215.54  227.11  4.29  11.57  
5  214.46  198.44  226.00  11.54  27.56  
6  200.93  191.31  224.75  23.82  33.44  
均值  10.80  17.10  
方差  7.23  11.63 
Fig.4 Fourier spectrum of subimages and the related Gaussian curves图4 子图像傅里叶谱及相应高斯曲线图 
Tab.2 Wind speed retrieval by CMOD4 model and CMOD5 model based on the wind direction by GaussianFFT表2 高斯曲线拟合FFT风向反演的CMOD4和CMOD5风速 
子图名  CCMP风向/°  CCMP风速值/(m/s)  高斯曲线拟合FFT风向/°  CMOD4风速值/(m/s)  CMOD4风速值与CCMP风速值差/(m/s)  CMOD5风速值/(m/s)  CMOD5风速值与CCMP风速值差/(m/s)  

1  227.28  11.29  218.32  14.46  3.17  19.25  7.96  
2  226.19  11.57  222.01  14.19  2.62  17.57  6.00  
3  224.99  11.77  213.00  15.85  4.08  18.53  6.76  
4  227.11  11.21  222.82  13.16  1.95  17.13  5.92  
5  226.00  11.44  214.46  14.15  2.71  18.05  6.61  
6  224.75  11.61  200.93  17.50  5.89  21.38  9.77  
均值  3.40  7.17  
方差  1.41  1.47 
Fig.5 Wind retrieval result of ASAR image图5 ASAR影像风场反演结果图 
The authors have declared that no competing interests exist.
[1] 

[2] 

[3] 

[4] 

[5] 

[6] 

[7] 

[8] 

[9] 

[10] 

[11] 

[12] 

[13] 
[

[14] 
[

[15] 

[16] 
[

[17] 

[18] 

[19] 

[20] 
[

[21] 
[

[22] 
[

[23] 

[24] 

[25] 

/
〈  〉 