The response of ENSO flavors to mid-Holocene climate: Implications for proxy interpretation

The response of El Nino-Southern Oscillation (ENSO) to mid-Holocene boundary conditions remains an open question: paleoclimate proxies and climate model simulations do not agree in the magnitude of the reduction of ENSO variability, while recent proxy evidence from fossil corals from the central Pacific show that the reduction in mid-Holocene ENSO variability compared to the end of the twentieth century is not different from the reduction during other Holocene time intervals. This is inconsistent with the interpretation of lake and ocean sediment records from the eastern Pacific, which show a significant reduction compared to all other Holocene periods. In order to reconcile the seemingly conflicting proxy evidence from the eastern and central Pacific, we hypothesize that ENSO remained active during the mid-Holocene; however, there was a change in the spatial pattern of the sea surface temperature anomalies, also known as ENSO flavors. Using National Center for Atmospheric Research's Community Climate System Model version 4 forced with mid-Holocene orbital conditions, we find that the frequency of occurrence of the strongest eastern Pacific (EP) events decreases in the mid-Holocene and their variance is reduced by approximate to 30%, while the frequency of central Pacific (CP) events slightly increases and their variance does not change. We also find a shift in the seasonality of EP events, but not in that of CP events. Lastly, mid-Holocene EP events develop more slowly and decay faster. The differential response of ENSO flavors to mid-Holocene forcing is remotely forced by the West Pacific, where a weakening of the trade winds in early boreal spring in the mid-Holocene initiates an anomalous downwelling annual Kelvin wave, which reaches the eastern Pacific during the ENSO development season, weakens the upper ocean stratification, and results in reduced ENSO upwelling feedback. The simulated reduction in the EP flavor versus the CP flavor in the mid-Holocene is consistent with proxy evidence: The teleconnection patterns of the two flavors with temperature, precipitation, and salinity are distinct, and proxies from different regions of the Pacific might be recording variability associated with only one of the flavors, or some combination of their relative effects.