Pacific warm pool and cold tongue evolution since late Miocene: Implications for El Niño

El Niño - Southern Oscillation (ENSO) is a consequence of tropical Pacific ocean-atmosphere dynamics and exerts control on Earth's interannual climate variability. Sea surface temperatures (SSTs) in the Pacific warm pool and cold tongue have been shown to be diagnostic for the strength of the Walker Circulation and ENSO conditions. Ancient SST reconstructions suggest that the temperature of the Pacific warm pool was invariant during the Plio-Pleistocene, whereas the cold tongue was much warmer in the Pliocene and then subsequently cooled. The appearance of a negligible east-west equatorial Pacific Ocean temperature gradient during the Pliocene is used to infer substantially reduced cold water equatorial upwelling, a weak Walker Circulation, and a "permanent" El Niño. Some recent records suggest otherwise. For example, high-resolution records in Pacific and Mediterranean region show quasi-periodic climate variability (2-7 years) during the late Miocene - Pliocene, consistent with ENSO conditions and in conflict with the prevailing climate paradigm. Here we show a multi-proxy (TEX86, Uk'37), multi-site reconstruction of SSTs in the Pacific warm pool (ODP Sites 806 and 1143) and cold tongue (ODP 846 and 850) since late Miocene (~12 Ma). TEX86 temperatures from both sites show long-term cooling of the western warm pool since late Miocene, in contrast to an apparent invariant temporal character previously assumed. At Site 850, cold tongue temperatures from both TEX86 and Uk'37 also show a cooling trend, with TEX86 much more variable than Uk'37. Further, TEX86-calculated temperatures are consistently cooler than Uk'37 temperatures since ~6 Ma, and suggest different production depths or seasons, whereas deviations of TEX86 and Uk'37 signals might be recording physical/chemical property changes in the cold tongue (e.g., the strength and seasonality of upwelling, stratification and nutrient supply). When all records are stacked, a ~3°C zonal temperature gradient persists between 12 and 6 Ma, and then increases toward modern-day conditions. Considering sedimentary records that show interannual variability during late Miocene - Pliocene, a ~3°C mean zonal gradient might be large enough to sustain El Niño - La Niña variability and thus directly challenges the notion of a permanent El Niño state during warmer global conditions. The persistence a significant and consistent zonal gradient during warmer Miocene conditions suggests that the Pliocene evolution of stronger E-W equatorial SST gradients was driven by non-linear effects related to declining global temperatures or resulted from threshold effects related to the tectonic evolution of the region.