EnsoDuration
Computes the number of month during which sea surface temperature anomalies (SSTA) lasts in the central equatorial Pacific. It is based on the central equatorial Pacific SSTA (horizontal Niño3.4 average) during 6 years centered on ENSO peak between model and observations
TropFlux 1979-2018 (main)
20CRv2 1871-2012, ERA-Interim 1979-2018, ERSSTv5 1854-2018, HadISST 1870-2018, NCEP2 1979-2018
Niño3.4
None
- seasonal cycle removed
- detrending (if applicable)
- smoothed with time running average
- spatial average
- December Niño3.4 SSTA regressed onto SSTA time series
- nbr of month during which ENSO life-cycle > 0.25
- abs((model-ref)/ref)*100
monthly
% of error
sea surface temperature (SST)
The first level shows the diagnostic used to compute the metric and highlight the difference between the model and the reference.
Figure 1: period during which ENSO life-cycle > 0.25, showing the "duration" of SSTA (usually close to the reference but slightly longer). The black and blue markers show respectively the reference and the model. The metric derived is the absolute value of the relative difference: abs((model-ref)/ref)*100.
The second level shows the temporal evolution of SSTA during ENSO: the ENSO life-cycle.
Figure 2: temporal structure of sea surface temperature anomalies (SSTA) in the central equatorial Pacific (Niño3.4 averaged), showing the temporal evolution of SSTA (usually too frequent swing from negative to positive SSTA). The black and blue curves show respectively the reference and the model..
The third level is similar to the first, but with La Niña and El Niño events.
Figure 3: distributions of durations of ENSO events, showing usually too long and too symmetrical durations (some La Niña events should last 2 years, but it is not the case for El Niño). Here the duration of La Niña is too small and 2-years La Niña are unlikely, while 2-years El Niño are simulated. In addition, simulated distributions are too narrow, showing that the model does not simulate the observed diversity of events duration. The black and blue boxplots show respectively the reference and the model, the left and right panels show respectively La Niña distributions and the El Niño distributions. Boxplots: whiskers extend to the 10th and 90th percentiles; boxes encompass the 25th and 75th percentiles; and a diamond marks the mean. La Niña (El Niño) events are defined when December Niño3.4 SSTA (December value smoothed with a 5-month triangular-weighted moving average) is below -0.75 (above 0.75) standard deviation.