@article{fdi:010082142,
  title = {{A} tropical {A}tlantic dynamics analysis by combining machine learning and satellite data},
  author = {{A}rnault, {S}abine and {T}hiria, {S}. and {C}repon, {M}. and {K}aly, {F}.},
  editor = {},
  language = {{ENG}},
  abstract = {{T}he western tropical {A}tlantic {O}cean is a very energetic and highly variable region. {I}t is one of the main contributors to the interhemispheric mass and heat transports. {T}his study aim is to give a new picture of the space and time variability of this region using statistical tools applied to five different satellite measurements ({S}ea {S}urface {T}emperature, {S}ea {S}urface {S}alinity, ocean topography, wind stress vectors). {W}e first processed each data set by using a {S}elf-{O}rganizing {M}aps ({SOM}), which is an efficient clustering methodology based on non-linear artificial neural networks to compress the information embedded in the data. {T}he {SOM} was then combined with a {H}ierarchical {A}scendant {C}lassification ({HAC}) to cluster the different phenomena in a small number of classes whose physical characteristics are easy to identify. {T}hree classes were identified which allowed us to analyse the dynamics of the {N}orth {B}razil {C}urrent, and the {N}orth {E}quatorial {C}ountercurrent, respectively, and their links with the {I}nter-{T}ropical {C}onvergence {Z}one and the {A}mazon and {O}rinoco river runoffs. {T}he {SOM} + {HAC} analysis gave a coherent picture of the concomitant seasonal variability of the variables. {F}urthermore, we were able to point out the correlations existing between salinity features recently discovered and wind, temperature, and dynamic topography structures. {A}pplying our method to the interannual signals, we showed a year to year variability which deserves further analysis.},
  keywords = {{A}tlantic ; {N}orth {B}razil {C}urrent ; {M}achine learning self-organizing map ; {S}atellite observations ; {ATLANTIQUE} ; {ZONE} {TROPICALE} ; {BRESIL} {NORD} {COURANT}},
  booktitle = {25 years of progress in radar altimetry},
  journal = {{A}dvances in {S}pace {R}esearch},
  volume = {68},
  numero = {2 {N}o {S}p{\'e}cial},
  pages = {467--486},
  ISSN = {0273-1177},
  year = {2021},
  DOI = {10.1016/j.asr.2020.09.044},
  URL = {https://www.documentation.ird.fr/hor/fdi:010082142},
}