Nonlinear Controls on the Persistence of La Nina

A large fraction (35%-50%) of observed La Nina events last two years or longer, in contrast to the great majority of El Nino events, which last one year. Here, the authors explore the nonlinear processes responsible for the multiyear persistence of La Nina in the Community Climate System Model, version 4 (CCSM4), a coupled climate model that simulates the asymmetric duration of La Nina and El Nino events realistically. The authors develop a nonlinear delayed-oscillator (NDO) model of the El Nino-Southern Oscillation (ENSO) to explore the mechanisms governing the duration of La Nina. The NDO includes nonlinear and seasonally dependent feedbacks derived from the CCSM4 heat budget, which allow it to simulate key ENSO features in quantitative agreement with CCSM4.Sensitivity experiments with the NDO show that the nonlinearity in the delayed thermocline feedback is the sole process controlling the duration of La Nina events. The authors' results show that, as La Nina events become stronger, the delayed thermocline response does not increase proportionally. This nonlinearity arises from two processes: 1) the response of winds to sea surface temperature anomalies and 2) the ability of thermocline depth anomalies to influence temperatures at the base of the mixed layer. Thus, strong La Nina events require that the thermocline remains deeper for longer than 1 yr for sea surface temperatures to return to neutral. Ocean reanalysis data show evidence for this thermocline nonlinearity, suggesting that this process could be at work in nature.