@article{PAR00026227,
  title = {{S}ubsurface eddy detection optimized with potential vorticity from models in the {A}rabian {S}ea},
  author = {{E}rnst, {P}. {A}. and {S}ubrahmanyam, {B}. and {M}orel, {Y}. and {T}rott, {C}. {B}. and {C}haigneau, {A}lexis},
  editor = {},
  language = {{ENG}},
  abstract = {{C}oherent ocean vortices, or eddies, are usually tracked on the surface of the ocean. {H}owever, tracking subsurface eddies is important for a complete understanding of deep ocean circulation. {I}n this study, we develop an algo-rithm designed for the detection of subsurface eddies in the {A}rabian {S}ea using {N}ucleus for {E}uropean {M}odelling of the {O}cean ({NEMO}) model simulations. {W}e optimize each parameter of our algorithm to achieve favorable results when compared with an algorithm using sea surface height ({SSH}). {W}hen compared to similar methods, we find that using the re -scaled isopycnal potential vorticity ({PV}) is best for subsurface eddy detection. {W}e proceed to demonstrate that our new al-gorithm can detect eddies successfully between specific isopycnals, such as those that define the {R}ed {S}ea {W}ater ({RSW}). {I}n doing so, we showcase how our method can be used to describe the properties of eddies within the {RSW} and even identify specific long-lived subsurface eddies. {W}e conduct one such case study by discerning the structure of a completely subsur-face {RSW} eddy near the {C}hagos {A}rchipelago using {L}agrangian particle tracking and {PV} diagnostics. {W}e conclude that our rescaled {PV} method is an efficient tool for investigating eddy dynamics within the ocean's interior, and publicly provide our optimization methodology as a way for other researchers to develop their own subsurface detection algorithms with optimized parameters for any spatiotemporal model domain.{SIGNIFICANCE} {STATEMENT}: {E}ddies are a key part of ocean circulation both at the surface and in the subsurface. {T}he purpose of our study was to design the first detection method comprehensively optimized for subsurface eddy detection from numerical simulations. {W}e demonstrate that potential vorticity ({PV}) is the best field to use when algo-rithmically tracking eddies in subsurface water masses, using our new method to identify and track eddies in the {R}ed {S}ea {W}ater ({RSW}). {A}dditionally, our method allows us to efficiently evaluate the dynamics of eddies through potential vorticity diagnostics, exemplified with a previously undescribed eddy near the {C}hagos {A}rchipelago. {O}ur methodology can be used by future researchers to study the eddy dynamics hidden within subsurface water masses around the world.},
  keywords = {{D}iapycnal mixing ; {E}ddies ; {I}nstability ; {P}otential vorticity ; {B}ottom currents ; bottom water ; {W}ater masses storage ; {OCEN} {INDIEN} ; {ARABIQUE} {MER}},
  booktitle = {},
  journal = {{J}ournal of {A}tmospheric and {O}ceanic {T}echnology},
  volume = {40},
  numero = {6},
  pages = {677--700},
  ISSN = {0739-0572},
  year = {2023},
  DOI = {10.1175/jtech-d-22-0121.1},
  URL = {https://www.documentation.ird.fr/hor/{PAR}00026227},
}