Where the winds clash: what is really triggering El Niño initiation?

The study addresses what triggers El Niño initiation within the ENSO cycle, focusing on the role of the Easterly/Westerly Wind Convergence Zone (EWCZ). ENSO redistributes heat zonally in the tropical Pacific, producing irregular SST oscillations with profound global socioeconomic impacts, including effects on agriculture, inflation, and conflict risk. Despite advances following super-El Niño events (1997–1998; 2015–2016), key issues in ENSO dynamics and prediction remain unresolved. Existing theories emphasize ocean subsurface heat recharge, thermocline depth variability, and equatorial Kelvin and Rossby waves driven by wind anomalies. The authors hypothesize that the longitude of the EWCZ acts as a critical, synthetic parameter controlling El Niño initiation and intensity: when the EWCZ shifts sufficiently eastward, the generated Kelvin waves can reach the eastern boundary and trigger El Niño, with severity increasing with further eastward extension. The work aims to quantify EWCZ zonal shifts, relate them to internal wave generation, deep atmospheric convection, and the Southern Oscillation Index (SOI), and establish a mechanistic threshold for El Niño onset.

Prior research has established the importance of equatorial wave dynamics and wind forcing in ENSO. The western Pacific Warm Pool (WP) hosts the warmest SSTs (≥28–29 °C) and deep convection, with its eastern edge and convergence zone shifting in phase with SOI. Studies have shown that sequences of westerly wind (WW) events in the western/central Pacific excite Kelvin (KW) and Rossby waves (RW), driving anomalous zonal advection and thermocline variability. However, the contributions of KW and RW can cancel in the western equatorial Pacific, with SST variability in the east primarily due to KW. The literature highlights the roles of WW intensity, location, and timing, as well as easterly winds (EW), in El Niño development, and suggests that WW events may be modulated by the ENSO state and large-scale pressure patterns rather than being purely stochastic. Modeling and observational studies (e.g., for 1997/98) documented that WW events lead to eastward WP displacement and shifts in deep atmospheric convection. Building on these, the paper formalizes the EWCZ longitude as a concise descriptor integrating WW intensity/location and EW anomalies to assess ENSO preconditioning.

Data and period: ERA-Interim surface winds over 130°E–80°W, 5°S–5°N for 1979-01-01 to 2019-08-31: monthly u, v, w and 6-hourly surface analyses averaged to daily u. SOI from the Climatic Research Unit (UEA). Sea surface height (SSH) from satellite altimetry (Copernicus SEALEVEL_GLO_PHY_L4_MY_008_047), 0.25° resolution, 15°S–15°N, 130°E–80°W, 1993-01-01 to 2019-12-31, resampled weekly with cubic spline. Temperature–Salinity fields from Copernicus armor-3d (MULTIOBS_GLO_PHY_TSUV_3D_MYNRT_015_012), 0.25°, 50 vertical levels, 15°S–15°N, 130°E–80°W, 1993-01-16 to 2018-12-16; seawater density computed with CSIRO MATLAB seawater library. EWCZ detection: For each date, the meridional mean (5°S–5°N) of the zonal wind (u) along 130°E–100°W was fit with a step function. The longitudinal domain was split into two subsegments of varying length; for each split, mean u over each segment defined a step function. The best representation minimized RMS error relative to the original meridional-mean u; the step transition longitude defined the EWCZ (wind convergence) longitude. Wind convergence context: Daily meridional-mean u averaged in 10°-wide zonal windows immediately west and east of the EWCZ; their difference characterized convergence. Internal wave and thermodynamic structure: Time–depth (time–z) sections of density and temperature anomalies were computed by averaging within a box centered on the EWCZ longitude and extending 10° zonally (±5°) and 4° meridionally. Kelvin wave extraction: Following Boulanger and Menkes, SSH anomalies were decomposed into equatorial modes as a linear superposition of freely propagating RWs and KWs; the KW amplitude (r0) was estimated assuming phase speed c = 2.5 m s−1 and parabolic-cylinder meridional structures. Diagnostics and relationships: The longitude of maximum vertical wind advection (proxy for deep atmospheric convection) was derived from ERA-Interim monthly vertical wind fields; correlations were computed between EWCZ longitude and SOI, and between convection longitude and SOI. Time–longitude plots of vertical wind, with EWCZ overlaid, and daily KW amplitude with EWCZ track, were analyzed to relate KW formation regions, propagation, and El Niño events (ONI classification). Threshold behavior relative to thermocline slope was examined using known thermocline structure west/east of ~170°E, and preliminary checks of 20 °C isotherm depth from the TAO array.

• Internal downwelling intrusions (50–250 m) of low-density, warm water form regularly beneath the EWCZ, evidencing generation of internal Rossby and Kelvin waves beneath the wind convergence.
• The EWCZ longitude is tightly anti-correlated with SOI (Pearson r ≈ −0.9 using 6-month median filtered series, 1979–2019), indicating that WW/EW spatial shifts are linked to the large-scale pressure gradient in the tropical Pacific.
• The longitude of maximum atmospheric vertical convection and SOI are also anti-correlated (r ≈ −0.6), and convection shifts occur in phase with EWCZ shifts.
• Zonal wind structure relative to EWCZ: west of the convergence, meridional-mean daily u fluctuates roughly between −4 and +3 m s−1; east of the convergence it remains westward between −5.5 and −0.5 m s−1. The east–west wind difference is almost always negative (convergent), except brief instances in May/June 1983, 1984, and 1998.
• Kelvin waves form beneath the EWCZ and propagate eastward; only when the EWCZ lies east of approximately 175°E do the generated Kelvin waves reach the eastern boundary and an El Niño event is recorded. The farther east the EWCZ extends into the central Pacific, the more severe the ensuing El Niño.
• Dynamical explanation: West of ~170°E, RW and KW effects on thermocline depth largely cancel; east of ~170°E, where the thermocline slopes, RWs refract poleward and cease to interact with equatorial KWs, allowing KWs to efficiently deepen the thermocline and warm the east, initiating El Niño.
• Case study: In 2014, despite a strong Kelvin wave crossing the basin, the EWCZ remained west of 175°E and no El Niño developed; in 2015 the EWCZ intruded far eastward and a super El Niño occurred.
• The EWCZ longitude serves as a compact metric summarizing WW intensity/location and EW anomalies; its eastward displacement beyond ~175°E identifies favorable preconditioning and predicts stronger events.

The findings support the hypothesis that the zonal position of the Easterly/Westerly Wind Convergence Zone is a primary control on El Niño initiation. By showing that Kelvin and Rossby waves are consistently generated at the EWCZ, but that only eastward EWCZ placements beyond ~175°E allow Kelvin waves to modify the eastern thermocline effectively, the study explains observed event dependence on the location of wind forcing. The mechanistic framework reconciles prior observations of KW–RW cancellation in the western Pacific with enhanced KW efficacy over a sloping thermocline east of ~170°E, where RWs exit the equatorial waveguide. The strong anti-correlations between EWCZ longitude, atmospheric convection longitude, and SOI demonstrate that westerly wind events and convergence shifts are not purely stochastic; rather, they are embedded within the tropical Pacific climate state. This integrated perspective, linking WW/EW competition, WP displacement, convection, sea-level pressure gradients, and EWCZ position, elucidates why El Niño emerges as a logical consequence when the convergence shifts sufficiently eastward. The 2014–2015 sequence illustrates the threshold behavior and predictive value of EWCZ longitude.

The study introduces the longitude of the Easterly/Westerly Wind Convergence Zone as a synthetic, predictive parameter for El Niño initiation and intensity. Internal downwelling waves are regularly generated beneath the EWCZ; when the EWCZ lies east of ~175°E, resulting Kelvin waves reach the eastern boundary, deepen the thermocline, and trigger El Niño, with severity increasing as the EWCZ intrudes farther east. The EWCZ and deep convection shift in phase with SOI, indicating that WW events are modulated by the tropical Pacific climate state. This framework explains the 2014 false alarm and the 2015 super El Niño and provides a physically grounded threshold for event onset. Under projected climate change—weakening trades, altered thermocline, reduced zonal SST gradients, and potentially more intense/frequent El Niño—the approach offers a pathway to assess ENSO regime shifts. Future work will estimate long-term changes in EWCZ position and isothermal slopes to explain evolving El Niño characteristics and improve predictability.