Calculation of Spaceborne SAR Image Heading and Its Impact on Wind Speed Retrieval

The spaceborne Synthetic Aperture Radar (SAR) is a broadly acknowledged technique to monitor the vast open ocean. However, a frequent confusion between the SAR image azimuth angle $ {\phi }_{\mathrm{I}\mathrm{M}\mathrm{G}} $ and the satellite flight direction $ {\phi }_{\mathrm{S}\mathrm{A}\mathrm{T}} $ exists in the data processing and geophysical applications, which influences for instance the accuracy of SAR wind inversions. An overview of $ {\phi }_{\mathrm{I}\mathrm{M}\mathrm{G}} $ derivation from the transformation of the spatial satellite coordinate systems is presented in this paper, along with a method for calculating $ {\phi }_{\mathrm{I}\mathrm{M}\mathrm{G}} $ based on the Ground Control Points (GCP). By analyzing a set of 0.7 million Sentinel-1 SAR wave mode images acquired in 2016, we quantify the systematic bias between $ {\phi }_{\mathrm{I}\mathrm{M}\mathrm{G}} $ and $ {\phi }_{\mathrm{S}\mathrm{A}\mathrm{T}} $ and find out that this bias is related to the incidence angle and orbital direction. It can be accurately fitted using a fifth-order polynomial relative to latitude. The misuse of $ {\phi }_{\mathrm{S}\mathrm{A}\mathrm{T}} $ is demonstrated to approximate $ {\phi }_{\mathrm{I}\mathrm{M}\mathrm{G}} $ can lead to nonnegligible wind retrieval errors. The spatial distribution of this wind error is non-uniform and differs between ascending and descending orbits, with a wind speed bias of ±0.5 m·s^–1. Results obtained in this study evidence that the precise estimate of $ {\phi }_{\mathrm{I}\mathrm{M}\mathrm{G}} $ are significant to various SAR applications in relevant studies. The proposed method here also holds practical reference for other satellite SAR missions.