Retrieval of 2-D Ocean Total Surface Current From Sentinel-1 Doppler Frequency Shift Using 2-D Kalman Filtering

Abstract

Synthetic aperture radar (SAR) has all-weather, all-day, large-area observation capability and can measure ocean total surface currents with Doppler frequency shift. However, conventional SAR can only obtain the radial velocity of the ocean total surface current. In this letter, we propose a method for inverting ocean total surface current vectors using a two-dimensional (2D) Kalman filter and a priori information about ocean current fields from Ocean Surface Current Analyses Real-time (OSCAR). We establish a complete filtering process including filter creation, filtering path selection and fusion of filtering results. We construct the Kalman filter according to the motion characteristics of ocean currents and transform the ocean surface current vector inversion problem into a filtering problem. The introduction of two-dimensional filtering can better adapt to the directionality of ocean currents. Furthermore, we design the filtering path to obtain smoother filtering results. In order to synthesize the filtering results of different filtering paths, we use the Covariance Intersection (CI) fusion method in distributed fusion to obtain the final ocean current vector inversion results. The accuracy of the inversion is verified by the measured data. The Root Mean Square Error (RMSE) of the proposed method with the mean Doppler error of 5.43Hz is 0.182 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m/s</i> in eastward speed and 0.141 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m/s</i> in northward speed. The reduction is 0.009 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m/s</i> and 0.029 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m/s</i> against OSCAR.