부산대학교 도서관

Much of the internal tide (IT) is coherent with tidal forcing, but interactions with time‐varying ocean flows result in an incoherent component of IT (IIT), which will complicate interpretation of data from the upcoming Surface Water Ocean Topography (SWOT) mission. Here, we demonstrate a simple scheme for estimation of these IIT. We use a short‐time harmonic analysis of outputs of a wind‐ and tidally forced ocean model to generate an ensemble of IT variations in an ocean patch. Principal Components Analysis is used to define the dominant spatial patterns, and combined with sinusoidal time variations at tidal frequencies to define spatio‐temporal basis functions that are used to fit sparse altimetry data available in short time windows. The scheme is illustrated using a 1‐yr Hybrid Coordinate Ocean Model (HYCOM) run off the Amazon Shelf. Tests with synthetic (SWOT) and real (nadir altimetry) data demonstrate that a significant fraction of IIT can be recovered with this procedure. Plain Language Summary: A key objective of the Surface Water Ocean Topography (SWOT) mission is the study of short wavelength (submesoscale) features in the ocean circulation. Success will depend on accurate corrections for other physical processes in the ocean, including internal waves generated by interaction of the ocean tides with bottom topography. Here, we describe a new approach to estimation of these internal tide wave fields, based on directly fitting the SWOT data. Statistical analysis of outputs of a numerical model of the ocean (which includes tides) are used to derive spatial patterns typical of the internal wave field for a specific patch of ocean, and these are then used in the estimation scheme. Tests with synthetic and real data demonstrate that the new approach can produce useful corrections of time‐variable internal‐tide wave fields. Key Points: We describe a new approach to estimation of time‐varying (incoherent) internal tides for correction of Surface Water Ocean Topography dataSpatial basis function for data fitting are derived from statistical analysis of tide and wind forced ocean modelsTests on synthetic and real data show that useful corrections can be derived even with very sparse sampling [ABSTRACT FROM AUTHOR]