Sea Surface Wind Retrieval from Synthetic Aperture Radar Data by Deep Convolutional Neural Networks

The sea surface wind at low and moderate speed (< 20 m/s) can be well derived from Synthetic Aperture Radar (SAR) data by geophysical model functions (CMOD5, XMOD, CMOD-X0, or others). However, wind retrieval bias is usually large at high wind speed (> 20 m/s). In this study, we explore to use a novel deep convolutional neural networks (DCNN) architecture, U-Net, to retrieve the sea surface wind at high speed. The Sentinel-1B SAR cross-polarized (HV and VH) Normalized Radar Cross Section (NRCS), SAR incidence angle and the ancillary Global Forecast System (GFS) model wind directions are used as the U-Net input data. The GFS model wind speed is the ground-truth data. The results suggest that the U-Net wind retrieval model has a capability to retrieve the high wind speed from SAR data with the physical model-based data. The accuracy of physical model data directly affects the U-Net model results. Meanwhile, U-Net model has a better continuity in the joint area of two strips.

[1]  Xiaofeng Li,et al.  Retrieving Hurricane Wind Speed From Dominant Wave Parameters , 2017, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[2]  Paris W. Vachon,et al.  C-Band Cross-Polarization Wind Speed Retrieval , 2011, IEEE Geoscience and Remote Sensing Letters.

[3]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Xiaofeng Li,et al.  On the use of Sentinel-1 cross-polarization imagery for wind speed retrieval , 2017, 2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).

[5]  William Perrie,et al.  A Hurricane Wind Speed Retrieval Model for C-Band RADARSAT-2 Cross-Polarization ScanSAR Images , 2017, IEEE Transactions on Geoscience and Remote Sensing.

[6]  Xiaofeng Li,et al.  Comparison of Ocean-Surface Winds Retrieved From QuikSCAT Scatterometer and Radarsat-1 SAR in Offshore Waters of the U.S. West Coast , 2011, IEEE Geoscience and Remote Sensing Letters.

[7]  Geoffrey E. Hinton,et al.  Rectified Linear Units Improve Restricted Boltzmann Machines , 2010, ICML.

[8]  Guoqing Zhou,et al.  Sea Surface Wind Retrievals from SIR-C/X-SAR Data: A Revisit , 2015, Remote. Sens..

[9]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[10]  Lanqing Huang,et al.  Technical Evaluation of Sentinel-1 IW Mode Cross-Pol Radar Backscattering from the Ocean Surface in Moderate Wind Condition , 2017, Remote. Sens..

[11]  William Perrie,et al.  A C-band Geophysical Model Function for Determining Coastal Wind Speed Using Synthetic Aperture Radar , 2018, 2018 Progress in Electromagnetics Research Symposium (PIERS-Toyama).

[12]  Yanfei Li,et al.  Wind speed forecasting method based on deep learning strategy using empirical wavelet transform, long short term memory neural network and Elman neural network , 2018 .

[13]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[14]  Fouad Badran,et al.  Neural network wind retrieval from ERS-1 scatterometer data , 2000, Neurocomputing.

[15]  H. Hersbach,et al.  An improved C-band scatterometer ocean geophysical model function: CMOD5 , 2007 .

[16]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[17]  Jingsong Yang,et al.  Development of a Gray-Level Co-Occurrence Matrix-Based Texture Orientation Estimation Method and Its Application in Sea Surface Wind Direction Retrieval From SAR Imagery , 2018, IEEE Transactions on Geoscience and Remote Sensing.

[18]  Biao Zhang,et al.  Wind retrieval from synthetic aperture radar - an overview , 2013 .