A new-generation internal tide model ZHAO30yr is constructed using 30 years of sea surface height measurements from 1993 to 2022 made by 15 satellite altimetry missions. The model contains 8 mode-1 constituents (M2, S2, N2, K2, K1, O1, P1 and Q1) and 4 mode-2 constituents (M2, S2, K1 and O1). These constituents are respectively extracted by the same mapping technique that consists of two rounds of plane wave analysis with a spatial bandpass filter in between. Their prerequisite wavelengths are calculated from climatological annual-mean hydrographic profiles in the World Ocean Atlas 2018. Model errors are estimated to be 0.5-1 mm for all semidiurnal and diurnal constituents. The model is evaluated using 1 year of independent altimetry data in 2023. The global internal tide field is decomposed into multiconstituent multimodal multidirectional internal waves. Multiwave decomposition reveals unprecedented new features previously masked by multiwave interference. Numerous long-range internal tidal beams are observed to be associated with notable topographic features. Semidiurnal beams off the Amazon shelf and diurnal beams in the Arabian Sea are studied in detail. The M2 and K1 internal tide constituents are examined for their multiwave decomposition and isolated beams on a global scale.

[JGR manuscript 2024JC021174R] Seventy-five days of sea surface height measurements made by the Surface Water and Ocean Topography (SWOT) mission from September 7 to November 21, 2023 are used to explore SWOT's capability of observing internal tides. Mode-1 M2 internal tides are mapped by our updated mapping technique. SWOT-75d represents a 75-day instantaneous model. Nadir-30y is constructed using 30 years of nadir altimetry data from 1993 to 2022 and represents a climate normal. The nadir altimetry data in 2023 are used for model evaluation. Despite its large errors, SWOT-75d reveals the basic features of the global mode-1 M2 internal tide field, and causes positive variance reduction in regions of strong internal tides. Nadir-30y performs better overall, but SWOT-75d performs better in the tropical South Atlantic Ocean, the central North Pacific Ocean, and the Melanesian region. Evaluation using seasonally subsetted altimetry data reveals that M2 internal tides have significant temporal variations. SWOT-75d performs the best in fall, because the model is constructed using data largely in fall. SWOT-75d has large phase anomalies, which are spatially smoothed and used to adjust the phases in Nadir-30y. The phase-adjusted model can better make internal tide correction for SWOT and its performance is improved by 20%. Our results demonstrate that (1) mode-1 M2 internal tides can be extracted from 75 days of SWOT data by our mapping technique, and (2) the instantaneous internal tide model can be used to improve internal tide correction for SWOT. 

Satellite altimetry sea surface height (SSH) measurements from 1993 to 2022 are used to show the strengthened mode-1 M2 internal tides in the past 30 years. Two mode-1 M2 internal tide models M9509 and M1019 are constructed using the data in 1995-2009 and 2010-2019, respectively. The results show that the global mean M2 internal tides strengthened by 6.6% in energy. However, the internal tide strengthening is spatially inhomogeneous. Significantly strengthened internal tides are observed in a number of regions including the Aleutian Ridge and the Madagascar-Mascarene region. Weakened internal tides are observed in the central Pacific. On a global average, M1019 leads M9509 by about 10o (20 minutes in time), suggesting that the propagation speeds of M2 internal tides increased. M9509 and M1019 are evaluated using independent altimetry data in 1993-1994 and 2020-2022. The results show that M9509 and M1019 perform better for the data in 1993-1994 and 2020-2022, respectively.

Two moorings deployed for 75 days in 2019 and long-term satellite altimetry data reveal a spatially complex and temporally variable internal tidal field at the SWOT Cal/Val site off central California due to the interference of multiple seasonally-variable sources. Coherent tides account for $\sim$45\% of the potential energy. The south mooring exhibits more energetic semidiurnal tides, while the north mooring displays stronger mode-1 M$_2$ with an amplitude of $\sim$5.1 mm. These findings from in situ observations align with the analysis of 27-year altimetry data. The altimetry results indicate that the complex internal tidal field is attributed to multiple sources. Mode-1 tides primarily originate from the Mendocino Ridge and the 36.5\textendash37.5$^\circ$N California continental slope, while mode-2 tides are generated by local seamounts and Monterey Bay. The generation and propagation of these tides are influenced by mesoscale eddies and seasonal stratification. Seasonality is evident for mode-1 waves from three directions. Southward components from the Mendocino Ridge consistently play a dominant role ($\sim$268 MW) yearlong. We observed the strongest eastward waves during the fall and spring seasons, generated remotely from the Hawaiian Ridge. Westward waves from the 36.5\textendash37.5$^\circ$N California continental slope are weakest during summer, while those from the Southern California Bight are weakest during spring. The highest variability of energy flux is found in the westward waves ($\pm 22\%$), while the lowest is in the southward waves ($\pm 13\%$). These findings emphasize the importance of incorporating the seasonality and spatial variability of internal tides for the SWOT internal tidal correction.