EquatorMix Remix: Assimilation of a Process Study Campaign for Estimating Pacific Upwelling and Mixing Physics

The Equatorial Undercurrent (EUC) brings cold waters near the surface, where they are transformed through mixing and air-sea fluxes then exported to higher latitudes. This upwelling is not accomplished by EUC advection alone; much of the connectivity to the surface occurs via turbulent transport processes. Pacific Upwelling and Mixing Physics (PUMP) is a proposed effort to diagnose these vertical transport processes in the east-central equatorial Pacific. To inform PUMP we can look to the EquatorMix project, which took a highly relevant dataset on and around the Equator at 140°W while a tropical instability wave (TIW) front was passing through in October 2012. The EquatorMix campaign involved intensive observation of the upper ocean and of the atmosphere boundary layer with unmanned aerial vehicles. We propose high-resolution assimilation of these EquatorMix observations in an ocean model to yield understanding of upwelling physics, modeling needs, and observing design for PUMP. The data assimilation tests the hypothesis that we understand the important physics and can simulate them.
The connectivity of the region makes it necessary to understand both the large-scale state as well as the local mixing conditions. We have developed a data-assimilating model to estimate the large-scale state (Verdy et al., 2017). The 1/3° resolution Tropical Pacific Ocean State Estimate (TPOSE.3) provides overlapping 4-month state estimates from 2010 to 2020 that balance mass, heat, and momentum, adjusting forcing and initial conditions so the model evolution best matches the observations. The estimates were cross-validated against the TAO mooring array and showed skill at timescales greater than 20 days, but higher-frequency variability poorly reproduced partly due to lack of observations. We have recently enhanced this state estimate resolution to 1/6° (TPOSE.6), sharpening the simulation of features such as TIWs.
We propose to estimate the ocean state and atmospheric fluxes during EquatorMix with a 1/24° Tropical Pacific Ocean State Estimate (TPOSE.24). Importantly, TPOSE.24 will have fine vertical resolution, hypothetically improving the representation of shear layers and short vertical scales observed in EquatorMix. The large-scale context will be provided by nesting and initializing with TPOSE.6. Our goal is reproducing the mixing structures and fluxes of heat, salt, and momentum observed by EquatorMix, improving the model resolution and parameterizations as needed. Where TPOSE.24 drifts from data constraints, we will diagnose the sources of error due to unresolved or poorly modeled processes and act to correct these errors. Where TPOSE.24 is consistent with the observations, we will use it to address PUMP goals: we will diagnose the vertical fluxes and determine the time and space scales on which they vary. Beyond addressing these questions directly, we also view EquatorMix as a pilot observation program, demonstrating what components of a PUMP process study could look like and the usefulness of observations like EquatorMix for constraining the ocean state and quantifying processes.