@article{fdi:010062534,
  title = {{D}oes sea surface temperature outside the tropical {P}acific contribute to enhanced {ENSO} predictability ?},
  author = {{D}ayan, {H}. and {V}ialard, {J}{\'e}r{\^o}me and {I}zumo, {T}akeshi and {L}engaigne, {M}atthieu},
  editor = {},
  language = {{ENG}},
  abstract = {{I}n this paper we seek to identify inter-annual sea surface temperature anomalies ({SSTA}) patterns outside the tropical {P}acific that may influence {E}l {N}io/{S}outhern {O}scillation ({ENSO}) through atmospheric teleconnections. {W}e assume that a linear {ENSO} hindcast based on tropical {P}acific warm water volume and {N}io3.4 {SSTA} indices captures tropical {P}acific intrinsic predictability inherent to recharge oscillator dynamics. {T}his simple hindcast model displays statistically significant skill at the 95 % confidence level at leads of up to seven seasons ahead of the {ENSO} peak. {O}ur results reveal that {ENSO}-independent equatorial wind stress anomalies only significantly improve the skill of that linear hindcast at the 95 % level in boreal spring and summer before the {ENSO} peak and in boreal fall, five seasons ahead of the {ENSO} peak. {A}t those seasons, the robust large-scale {SST} patterns that provide a statistically significant enhancement of {ENSO} predictability are related to the {A}tlantic meridional mode and south {P}acific subtropical dipole mode in spring, the {I}ndian {O}cean {D}ipole and the south {A}tlantic subtropical dipole mode in fall. {W}hile the first two regions display significant simultaneous correlations with western equatorial {P}acific wind stress in three reanalyses ({ERA}-{I}, {NCEP} and {NCEP}2), the {I}ndian {O}cean {D}ipole and south {A}tlantic subtropical dipole mode correlation with {P}acific winds is less robust amongst re-analyses. {W}e discuss our results in view of other studies that suggest a remote influence of various regions on {ENSO}. {A}lthough modest, the sensitivity of our results to the dataset and to details of the analysis method illustrates that finding regions that influence {ENSO} from the statistical analysis of observations is a difficult task.},
  keywords = {{E}l {N}ino {S}outhern {O}scillation ({ENSO}) ; {P}redictability ; {T}eleconnections ; {PACIFIQUE} ; {ZONE} {TROPICALE}},
  booktitle = {},
  journal = {{C}limate {D}ynamics},
  volume = {43},
  numero = {5-6},
  pages = {1311--1325},
  ISSN = {0930-7575},
  year = {2014},
  DOI = {10.1007/s00382-013-1946-y},
  URL = {https://www.documentation.ird.fr/hor/fdi:010062534},
}