From the analysis of the isopycnal surface and my readings, especially of Wiliams and Follows (2003), Lee *et al.* (2007) and Nishikawa *et al.* (2010), this is what I need to do:

- computation over multiple years of the mean mass streamfunction and its eddy component to compare with Nishikawa
*et al.*(2010): Is my calculation right? Do we have diapycnal transport? - calculation of the mean entrainment velocity to compare with Nishikawa
*et al.*(2010): Is the present OFES model output enough to compute this quantity? If yes, apply to other (shallower) isopycnals to close the time-mean budget of isopycnic nutrient. (question: can we estimate the entrainment velocity from the mean mass streamfunction?) - computation of the time-mean and zonal-mean isopycnic nutrient and of the mean advection, eddy advection and eddy diffusion of nutrients (see Lee
*et al*. 2007 on how to interpret the direction of the eddy advection and diffusion given the eddy component of the mean mass streamfunction and the time-mean and zonally-averaged isopycnic nutrient, and the relative slopes of isopycnals and nutrient contours): Can we describe the role of mean flow and eddies in maintaining the nutricline in the subtropical gyre? - mean nutrient budget for a cell following the isopycnal layer: local time rate of change, mean + eddy isopycnal advection, mean + eddy isopycnal diffusion, mean + eddy diapycnal advection, mean + eddy diapycnal diffusion, the rest being consumption and reminelarization.
- perform correlation between shallowing events, ratio of relative vorticity to Coriolis parameter (needs to be high for frontogenesis to occur), vertical isopycnal velocity, nutrient concentration and its time rate of change, potential vorticity