The paper presents a precise experimental design to sample the ocean with a quasi Lagrangian approach for biogeochemical studies. It analyses in detail the success of such an approach in a specific region, with a rather low eddy kinetic energy, using relevant tools. Nevertheless, there is perhaps a last point, which should deserve some attention.
The implicit basic hypothesis or simplification of these biogeochemical cruises assumes that the ocean is mostly 1D at first order. This analysis analyses the variability of the data at local scale (relative to the initial conditions) considering the different layers are decoupled : the relative variability is quantified for each depth or layer independently. Nevertheless, the vertical shear (its impact can be seen through the different trajectories in fig 15) makes the system only quasi Lagrangien, as pointed out by the authors. Nevertheless, the different biological stations, and their time evolution within several days, are generally interpreted as Lagrangian on the whole depth (the traps collect material coming from the vertical during several days): how the proposed analysis can support that hypothesis? Something should be said about that issue, certainly related to the quasi homogeneity of the sampled area, the time scales of the studied processes which could create and modify the biogeochemical fluxes, etc…. More generally, this work uses only T – S parameters. Something about biogeochemistry, which is the core of OUTPACE, should be presented in the conclusion/discussion part.
Moreover, most of the data used in this study (CTD, …) sample the very surface, down to 200m, whereas there are sediment traps at 250 and 500m. The work is build on analyses on isopycnal surfaces, but what happens below 200 m in terms of variability ? The only information (on dispersion) is implicitly given in figure 15 with the SADCP trajectories (concerning that point, perhaps, at least for LDA, different colors could be used for different depths to better visualize this verical shear).