On the Intraseasonal Oceanic Processes Constrained by Data Assimilation: A Case Study of the Tropical Pacific

Abstract This study investigates the ability of a global ocean reanalysis at 1/12° horizontal resolution, GLORYS12, to represent oceanic processes at intraseasonal and higher-frequency scales. GLORYS12, which includes data assimilation of satellite and multi-instrument in situ observations, is compared to a twin-free simulation (with no assimilation) in the tropical Pacific Ocean. Spectral analyses show that data assimilation improves the realism of sea surface height intraseasonal variability in the entire tropical Pacific Ocean, in both amplitude and phase, with an increase in the amplitude of more than 50% for the 20–90-day band and up to 15% for the 2–20-day band. The improvement is largest along the 5°N/S latitudes, where the magnitude of tropical instability waves is maximum, but is limited along the equator where steric height variability is dominated by intraseasonal oceanic Kelvin waves, already well represented in the free simulation. Wavenumber–frequency spectra show that data assimilation constraint improves both the spatial and temporal scales of intraseasonal waves and their timing. Data assimilation impacts the realism of oceanic simulations in two ways. By modifying the background oceanic stratification, it corrects the phase speed of westward-propagating waves. It is also shown that the intraseasonal component of analysis increments (data assimilation corrections applied) is dynamically consistent and exhibits clear intraseasonal propagation. By demonstrating the benefits of data assimilation for intraseasonal processes in the tropical Pacific Ocean, this study highlights the high value of both in situ and satellite observations to constrain ocean models in a wide range of time scales.