Fig. 5: Relationship between hypothesized ENSO-sensitive background conditions and measures of ENSO activity.

Records of ENSO-sensitive background conditions are in the left column. Sample intervals are shown as vertical gray bars with widths proportional to the 1σ age uncertainty. Middle column shows the relationship between the climate background condition and the mean of the 94^th–98^th quantiles from our sample intervals. Right column shows the relationship between the climate background conditions and median absolute deviation (MAD) of calculated temperatures. In the right two columns vertical gray bars represent the ~1-sigma uncertainty in the mean of the 94^th–98^th quantile (middle) and standard error of the MAD (right). Gray horizontal bars depict the age-related uncertainty in the climate background conditions. For the correlation plots, the mid-Holocene interval is shown as an X, Marine Isotope Stage (MIS) 2–3 intervals are squares, MIS 5 intervals are circles, MIS 6 intervals are triangles, MIS 7 intervals are diamonds, and MIS 8 intervals are inverted triangles as in Fig. 3. Glacial intervals are depicted in blue, interglacial intervals are in orange. Filled markers represent intervals with altered El Niño amplitude significant at the 90^% confidence level (black) or increased El Niño amplitude (gray). Correlation coefficients (R value) and statistical significance from a linear fit model (LM) and weighted bivariate linear regression (WLR) between each climate parameter and the measures of ENSO activity are shown. a June insolation at 0°N³¹. b Multi-proxy tropical Pacific East-West SST gradient³³; c Inferred position of the intertropical convergence zone (ITCZ) from Titanium Mass Accumulation Rate (MAR)³²; d Vertical contrast (inversely related to mixed layer depth) from ODP Site 851 in the Eastern Pacific. Vertical contrast is calculated as the difference between G. sacculifer and Globorotalia tumida δ¹⁸O³⁶.