Modeling the stable water isotope expression of El Niño in the Pliocene: Implications for the interpretation of proxy data

The El Niño–Southern Oscillation (ENSO) drives interannual climate variability; hence, its behavior over a range of climates needs to be understood. It is therefore important to verify that the paleoarchives, used for preinstrumental ENSO studies, can accurately record ENSO signals. Here we use the isotope enabled Hadley Centre General Circulation Model, HadCM3, to investigate ENSO signals in paleoarchives from a warm paleoclimate, the mid-Pliocene Warm Period (mPWP: 3.3–3.0 Ma). Continuous (e.g., coral) and discrete (e.g., foraminifera) proxy data are simulated throughout the tropical Pacific, and ENSO events suggested by the pseudoproxy data are assessed using modeled ENSO indices. HadCM3 suggests that the ability to reconstruct ENSO from coral data is predominantly dependent on location. However, since modeled ENSO is slightly stronger in the mPWP than the preindustrial, ENSO is slightly easier to detect in mPWP aged coral. HadCM3 also suggests that using statistics from a number of individual foraminifera (individual foraminifera analysis, IFA) generally provides more accurate ENSO information for the mPWP than for the preindustrial, particularly in the western and central Pacific. However, a test case from the eastern Pacific showed that for some locations, the IFA method can work well for the preindustrial but be unreliable for a different climate. The work highlights that sites used for paleo-ENSO analysis should be chosen with extreme care in order to avoid unreliable results. Although a site with good skill for preindustrial ENSO will usually have good skill for assessing mPWP ENSO, this is not always the case.