Accurately Extraction of Ocean Tidal Constituents from Coastal Satellite Altimeter Records

Yanguang Fu¹Fukai Peng²Yikai Feng^1*Mehdi Khaki³Xiaolong Mi⁴

• ¹First Institute of Oceanography, Ministry of Natural Resources, Qingdao, China
• ²School of Geospatial Engineering and Science, Sun Yat-Sen University, Guangzhou 510275, guangzhou, China
• ³School of Engineering, The University of Newcastle, Callaghan, NSW 2308, Australia, Newcastle, Australia
• ⁴Department of Land Surveying and Geo-Informatics, Faculty of Construction and Environment, Hong Kong Polytechnic University, Kowloon, Hong Kong, SAR China

Extracting tidal constituents in coastal regions remains a significant challenge due to complex bathymetry, nonlinear shallow-water effects, and land contamination in satellite altimetry measurements. While tide gauge records provide high-precision tidal observations, their sparse spatial coverage limits their applicability for global coastal tidal studies. Similarly, global tidal models, despite incorporating data assimilation, often exhibit substantial discrepancies in coastal areas due to their limited spatial resolution and the lack of nearshore observational constraints. To address these limitations, this study utilizes the newly released International Altimetry Service 2024 (IAS2024) dataset, derived from reprocessed Jason-1/2/3 satellite altimetry data spanning 2002-2022, to extract ten primary tidal constituents (Q1, O1, P1, K1, N2, M2, S2, K2, Sa, and Ssa) in global coastal waters. The accuracy of these tidal extractions is evaluated through a comparative analysis with four state-of-the-art global tidal models (DTU16, EOT20, FES2014, and FES2022) and 164 tide gauge records. Results demonstrate that IAS2024 achieves accuracy levels comparable to EOT20 and superior to FES2014 and FES2022, with performance approaching that of DTU16-an empirical model assimilating extensive tide gauge and satellite altimetry data. For the eight major tidal constituents, IAS2024 exhibits a root sum square error of 11.26 cm, closely aligning with DTU16 (11.23 cm), EOT20 (11.68 cm), and FES2022 (11.26 cm). In comparisons with tide gauge records, the relative errors of IAS2024 are 14.16% for O1, 16.6% for M2, 15.4% for K1, and 17.7% for S2, demonstrating precision levels consistent with leading global tidal models. Notably, IAS2024 significantly outperforms traditional tidal models in resolving long-period tidal constituents (Sa and Ssa), achieving amplitude correlation coefficients of 0.924 (Sa) and 0.701 (Ssa), markedly surpassing EOT20 (0.890, 0.006) and FES2022 (0.457, 0.192). Furthermore, IAS2024 maintains high accuracy within 10 km of the coast, where conventional satellite altimetry data typically experience severe degradation. The dataset's ability to resolve long-period tidal variations also enhances its applicability to coastal sea level research, tidal energy assessments, and hydrodynamic modeling. This study highlights the significant advantages of IAS2024 in coastal tidal extraction and provides new insights into nearshore tidal dynamics, offering a valuable dataset for advancing global and regional tidal studies.