Sea Surface Height Estimation with Multi-GNSS and Wavelet De-noising

Fade Chen,Fei Guo,Lilong Liu

doi:10.1038/s41598-019-51802-9

Fade Chen, Fei Guo + Show 1 more

Open Access

https://doi.org/10.1038/s41598-019-51802-9

Copy DOI

Abstract

This paper presents a new sea surface height (SSH) estimation using GNSS reflectometry (GNSS-R). It is a cost-effective remote sensing technique and owns long-term stability besides high temporal and spatial resolution. Initial in-situ SSH estimates are first produced by using the SNR data of BDS (L1, L5, L7), GPS (L1, L2, L5), and GLONASS (L1, L2), of MAYG station, which is located in Mayotte, France near the Indian Ocean. The results of observation data over a period of seven days showed that the root mean square error (RMSE) of SSH estimation is about 32 cm and the correlation coefficient is about 0.83. The tidal waveform is reconstructed based on the initial SSH estimates by utilizing the wavelet de-noising technique. By comparing the tide gauge measurements with the reconstructed tidal waveform at SSH estimation instants, the SSH estimation errors can be obtained. The results demonstrate that the correlation coefficient and RMSE of the wavelet de-noising based SSH estimation is 0.95 and 19 cm, respectively. Compared with the initial estimation results, the correlation coefficient is improved by about 14.5%, while the RMSE is reduced by 40.6%.

Highlights

Obtaining accurate Sea Surface Height (SSH) is significant for human living, especially for those live along ocean coasts
There is a good agreement between tide gauge sea surface height (SSH) observations and SSH estimates obtained by the Global Navigation Satellite System (GNSS)-R methods over a period of seven days
The initial in-situ SSH estimates were obtained through Lomb-Scargle periodogram (LSP) spectral analysis on the detrended Signal-to-Noise Ratio (SNR) time series

Summary

Introduction

Obtaining accurate Sea Surface Height (SSH) is significant for human living, especially for those live along ocean coasts. The reflected signals of Global Navigation Satellite System (GNSS) are used to retrieve a range of geophysical parameters such as SSH, soil moisture, ocean wind speed, etc.[10,11,12,13,14,15,16,17,18]. Santamaría-Gómez et al developed an approach to extract SNR data dominated by sea-surface reflections and to remove SNR frequency changes caused by the dynamic sea surface[23] They successfully demonstrated its ability to estimate local SSH and improved accuracy. The focus is on GNSS-R ocean surface altimetry based on the SNR data of multiple satellite constellations and multiple frequencies, namely BDS (L1, L5, L7), GPS (L1, L2, L5) and GLONASS (L1, L2).

Methods

Results

Conclusion