The paper identifies TRAPs – objectively defined (frame-independent) analogs of hyperbolic points, in the eastern N. Pacific using a 20-year long record of mesoscale surface currents derived from SSH altimetry and wind-induced Ekman estimates. The authors present statistics of TRAPS – lifetime, propagation speed, strength of attraction, probability of occurrence etc., investigate the relative vorticity patterns around TRAPs, and look for the influence of TRAPs on real drifters passing nearby by constructing composite maps of drifter velocities in the vicinity of TRAPs.  

The results suggest that TRAPs are most commonly located in regions between mesoscale eddies, have similar westward propagation speeds to mesoscale eddies, and are associated with the hyperbolic flow geometry in their vicinity. The most typical vorticity patterns around TRAPs correspond to a quadrupole – 2 cyclonic and 2 anticyclonic vortices, with a TRAP at the center between them.

The paper is easy to read and the statistics is carefully estimated and clearly presented.

My main comments are:

• The mesoscale currents used for the analysis are likely only marginally applicable for predicting the small-scale distribution of garbage on a day-to-day basis and thus would only marginally help in any real cleanup efforts, which is the main motivation that the authors give for this work. I am not convinced that TRAPs and their statistics would be unchanged if the submesoscale flow features were resolved. This should be clearly explained, so as not to overstate the usability of TRAPs for real garbage cleanup.

• There are several subjective choices that likely affect the statistical results. First, limiting traps to 1 deg arclength seems both arbitrary and unnecessary. I don’t think it is necessary to put any limit on TRAPs lengths. It would be better to identify the full extent of a TRAP line length and analyze its statistics. Second, a 75 km radius was used for the pairing of drifters and TRAPs. Again, this seems like an arbitrary and unnecessary choice. It would be better to define the area of influence for each TRAP around the local minimum, perform statistical analysis of the basin of influence, and use the basin of influence for pairing with drifters.
• The authors seem to suggest that drifters (and other floating objects like garbage) are more likely to be found near TRAPs than in other regions of the flow. However, I don’t think there is any confirmation of this claim in the paper. It would be interesting to compare the statistics for the drifters occurrences and retention times within TRAPs to that within the mesoscale eddies and maybe even for a random subset of regions of comparable size and spatial distribution.

Please find my review in the attached document.