Interannual variability in the North Pacific Ocean from observations and a data-assimilating model

Interannual variability of the volume, heat, and freshwater circulation in the North Pacific Ocean is explored through a joint analysis of observations and the output from a data-assimilating model. High-resolution repeated expendable bathythermograph (XBT) transects provide an observational basis for analysis of transport of volume, heat, and freshwater in the North Pacific. The Estimating the Climate and Circulation of the Ocean (ECCO) Consortium uses the adjoint method to constrain an ocean circulation model with observations, producing dynamically consistent time-varying ocean state estimates. These state estimates provide a context in which the detailed information from the observations can be used for analysis of the mean and variability of ocean circulation.An initial analysis of volume transport in the Northeast Pacific demonstrates that comparisons between a global ocean state estimate and the data are useful in understanding the large-scale gyre interactions, as well as connections with larger-scale signals. To improve the accuracy of the ocean state estimate in the North Pacific, several experiments are performed with the ECCO model in a regional setting. First, we withhold subsets of the data from the assimilation to emphasize the importance of including all available data in order to obtain an accurate state estimate. Separately, we determine that increasing the weights on the subsurface data increases the accuracy of the subsurface estimate with minimal cost to the accuracy of the surface estimate. This new North Pacific state estimate is used to develop heat and freshwater budgets. A trans-Pacific XBT track defines the southern boundary of a closed region, and in that region the balance between cross-track advective transport and surface fluxes gives an estimate of the time-varying storage of heat and freshwater. The mean estimates of transport and storage compare well with previous research. In addition, estimates of the magnitude of variability are provided. The freshwater budget is found to be relatively stable, while the heat budget has large interannual variability. Connections between the variability of the heat storage in the North Pacific and the El Nino/Southern Oscillation climate signal are found.