The high-resolution expendable bathythermograph (HR-XBT) network measures temperature down to approximately 800 m along fixed transects. In comparison, Argo floats are distributed throughout the global ocean and measure temperature and salinity down to approximately 2000 m. Over the 2004-2019 period, the HR-XBT network tended to have a greater sampling density near the coast, while Argo sampling density was generally equivalent to, or greater than, the HR-XBT network in the ocean interior. To take advantage of the benefits of each of these observing systems, a method for combining measurements from HR-XBT, Argo, and satellite altimetry observations was implemented. This method produced estimates of geostrophic velocity and transport in the upper 800 m normal to the HR-XBT transects in the Indian and Pacific Oceans over an approximately 16 year period at high spatial (0.1° along-transect spacing) and temporal (1 month) resolutions. The combined method better resolved the mean geostrophic velocity and transport in the western boundary currents and their recirculations compared to estimates from a 2004-2018 mean high-resolution climatology computed using only Argo data. An additional benefit of the combined method is that it provides a monthly time series of velocity and transport for each HR-XBT transect. This monthly time series captures the temporal variability and will allow for examination of possible drivers of ocean transport.

Abstract: Float trajectories are simulated using Lagrangian particle tracking software and eddy‐permitting ocean model output from the Estimating the Circulation and Climate of the Ocean, Phase II (ECCO2) project. We find that Argo‐like particles near strong mean flows tend to accelerate while at their parking depth. This effect is pronounced in western boundary current regions and in the Antarctic Circumpolar Current system. The acceleration is associated with eddy‐mean flow interactions: Eddies converge particles toward regions with stronger mean currents. Particles do not accelerate when they are advected by the eddy or mean flow alone. During a 9‐day parking period, speed increases induced by the eddy‐mean flow interactions can be as large as 2 cm s−1, representing roughly 10% of the mean velocity. If unaccounted for, this acceleration could bias velocities inferred from observed Argo float trajectories.

Abstract: Argo float trajectories are simulated in the southwest Pacific basin (25°–45°S, 170°E–165°W) using velocity fields from a 1/12° Southern Ocean model and a Lagrangian particle tracking model programmed to represent the vertical motions of profiling Argo floats. The system is applied to simulate both core Argo floats (typically parked at 1000-m depth and profiling to 2000-m depth) and Deep Argo floats (parked 500 m above the seafloor). The goal is to estimate probability density functions (PDFs) predicting future float positions. Differences are expected in the trajectory statistics, largely because of limitations in the temporal and spatial resolution of the model fields and uncertainties associated with a random walk component included in the particle advection scheme to represent this unresolved variability. Nonetheless, the core Argo float displacements over ~100-day time intervals are mostly consistent with the derived PDFs, particularly in regions with stable midlayer flows. For the Deep Argo floats, which are released into the open ocean and parked near the bottom, the simulations predict an average total displacement of less than 50 km within 100 days, in good agreement with the Deep Argo floats deployed as part of a pilot study. The study explores both the representativeness and the predictability of float displacements, with an aim to contribute to planning for the float observing system.