Upper ocean state estimation in the Southern Ocean Gas Exchange Experiment region using the four-dimensional variational technique

Abstract

A prototypical ocean circulation model, and associated 4‐D variational data assimilation system, is configured in support of the Southern Ocean Gas Exchange Experiment (Southern Ocean GasEx). The regional circulation model has 4 km horizontal resolution and 5 m vertical resolution in the upper ocean. In situ temperature and salinity observations and remotely sensed, gridded sea surface temperature and sea surface height observations are combined with the circulation model between 6 March and 4 April 2008. The benefits of customized data assimilation in support of a single oceanographic cruise are illustrated using Southern Ocean GasEx as a case study. The assimilated fields fit the observations within their respective uncertainties and fit 4–6 times better than the initial guess solution before assimilation. The fit of the assimilated fields is up to 100 times closer to observations than various global state estimate products using a cost function variance metric. The space‐time variability in mixed layer depth is consistent with the Southern Ocean GasEx observations and shows a deepening through March 2008. This variability is correlated with sea level height anomalies, sea surface temperature, and air temperature, during the assimilation window. The mixed layer depth estimate between 21 March and 4 April 2008 at the location of the Southern Ocean GasEx deliberate tracer release is 41–54 ± 6 m. The average zonal volume flux estimate is 16 ± 1.9 Sv (1 Sv is 106 m3 s−1) for latitudes 50.7°S–51.8°S, longitude 38.5°W. The zonal volume flux approximately halves during 6–16 March; temporal changes in volume flux are associated with geostrophic circulation but not with the Ekman circulation.

Southern Ocean Gas Exchange Experiment: Setting the stage

“… Del Castillo and Miller [2011] examine the colored dissolved organic matter dynamics during SO GasEx. Dwivedi et al [2011] use an ocean circulation model along with data assimilation of in situ and remote sensing data to simulate the ocean state during SO GasEx. V. P. Lance et al (Primary productivity, new productivity and carbon export during two Southern Ocean Gas Exchange (SO GasEx) tracer experiments, unpublished manuscript, 2011), Lee et al [2011], and R. C. Hamme et al (Dissolved O2/Ar and other methods reveal rapid changes in productivity during a Lagrangian experiment in the Southern Ocean, submitted to Journal of Geophysical Research , 2011) quantify productivity during SO GasEx.…”

Estimating the distribution of colored dissolved organic matter during the Southern Ocean Gas Exchange Experiment using four‐dimensional variational data assimilation

“…In (1), x ( t ) is the model state vector at time step t ; y ( t ) is the vector of observations; E ( t ) x ( t ) is the corresponding model counterpart; and is the background estimate of the model state at the initial time. R here is the error covariance matrix for the data‐model misfit terms, and P 0 is the background error covariance matrix for initial state [ Dwivedi et al ., ]. The assimilation process controls the initial temperature and salinity, initial horizontal velocity, initial SF 6 concentration, and initial CDOM absorption coefficients at 350, 380, and 400 nm.…”

“…Dwivedi et al . [] also use these data and only a brief description is given here. The biochemical measurements (in situ and satellite CDOM and in situ SF 6 ) were new to this state estimate.…”

“…We chose a constant value of 1°C for SST uncertainty as in Dwivedi et al . []. There were 18,928 SST observational data in the state estimate.…”

“…This study takes advantage of previous efforts by Del Castillo and Miller [], Dwivedi et al . [], and Ho et al . [].…”