This is the first analysis of the shallow and deep events for a set of isopycnal surfaces in OFES for the year 2004. The domain of study is the same as for the ARGO float: 210-220E and 20-25N.

Fig. 1 shows the depths of each isopycnal for the subdomain 214-216E and 22-23N and their respective spatial and time mean. As in the ARGO observations (*Fig. 1 of this note*), the standard deviation of the depth anomalies has a minimum near 200 m depth.

Fig. 2 shows the nitrate concentration (N) for the same subdomain and for each isopycnal surface. At depth, the standard devation is small suggesting that N follows potential density there. The standard devation is maximum within the nutricline (100-200 m depth). The distribution of N for each surface is highly skewed; that means the high N values observed are rare. Indeed, animation of N along 24.5 kg/m^3 shows that the high values occured during one single event. This event corresponds to the passage during July of a large deep nitrate event (Fig. 6). This event does not have a signature in SSHA (Fig. 5a) and the nitrate was a maximum trapped inside a cyclonic subsurface eddy (Fig. 5b).

What is interesting is the absence, at least in this subdomain, of large negative N anomalies. At depth, below the euphotic zone (like along 26 kg/m^3), this means that any coherent vortices are bringing water that has a larger nitrate concentration than the local region, consistent with the idea that the subtropical gyre is surrounded by region of high N to the north and to the south.

Figs. 3 and 4 shows the depth anomaly (DA) *versus* SSHA for σ = 24.5 kg/m^3 and 26.5 kg/m^3 respectively for the subdomain 214-216E and 22-23N. Note that all large negative DA for σ = 24.5 kg/m^3 and positive DA for σ = 26.5 kg/m^3 that occur with negative weak SSHA occur during the July event. Fig. 5 shows together SSHA and DA for surfaces at the center of the subdomain (215E, 22.5N). Notice that only in few cases, the DA and SSHA covary and not even for the largest events (like the July’s one).

A set of animation is available:

- animation of SSHA (SSHA is defined as SSH minus the 2004 time-mean SSH). I am worry about the jump in SSHA that we observe every 3 days.
- animation of DA for 24.5 and 26.5 kg/m^3
- animation of N along 24.5 kg/m^3