A Wind Speed Retrieval Model for Sentinel-1A EW Mode Cross-Polarization Images

Yuan Gao,Jian Sun,Changlong Guan,Lian Xie

doi:10.3390/rs11020153

Abstract

In contrast to co-polarization (VV or HH) synthetic aperture radar (SAR) images, cross-polarization (CP for VH or HV) SAR images can be used to retrieve sea surface wind speeds larger than 20 m/s without knowing the wind directions. In this paper, a new wind speed retrieval model is proposed for European Space Agency (ESA) Sentinel-1A (S-1A) Extra-Wide swath (EW) mode VH-polarized images. Nineteen S-1A images under tropical cyclone condition observed in the 2016 hurricane season and the matching data from the Soil Moisture Active Passive (SMAP) radiometer are collected and divided into two datasets. The relationships between normalized radar cross-section (NRCS), sea surface wind speed, wind direction and radar incidence angle are analyzed for each sub-band, and an empirical retrieval model is presented. To correct the large biases at the center and at the boundaries of each sub-band, a corrected model with an incidence angle factor is proposed. The new model is validated by comparing the wind speeds retrieved from S-1A images with the wind speeds measured by SMAP. The results suggest that the proposed model can be used to retrieve wind speeds up to 35 m/s for sub-bands 1 to 4 and 25 m/s for sub-band 5.

Highlights

A large number of geophysical model functions (GMF) have been presented to retrieve wind speeds from co-polarization (VV or HH) synthetic aperture radar (SAR) images
The aim of this study is to develop a new wind speed retrieval model for S-1A Extra-Wide swath (EW) mode VH-polarized images according to the relationships between noise-free normalized radar cross section (NRCS), sea surface wind speed and radar incidence angle
A new model is developed for retrieving sea surface wind speed from S-1A EW mode VH-polarized images. 19 noise-free S-1A images and matching data from Soil Moisture Active Passive (SMAP) radiometer under tropical cyclone conditions are collected and analyzed

Summary

Introduction

A large number of geophysical model functions (GMF) have been presented to retrieve wind speeds from co-polarization (VV or HH) SAR images. According to many C-band VV-polarized GMF models, the normalized radar cross section (NRCS) is dependent upon the wind speed at 10-m height, wind direction and radar incidence angle. Wind speed retrieval from co-polarization SAR images is known to have a number of limitations. The difficulty to obtain a collocated high-resolution wind direction field often leads to a decrease in the accuracy of wind speed retrieval [3,4,5,6]. The co-polarization NRCS is dampened at certain incidence angles, leading to a wind speed ambiguity problem [7]

Objectives

Findings

Discussion

Conclusion