A Modal Rendition of ENSO Diversity

Abstract

The El Nino and Southern Oscillation (ENSO) 'diversity' has been considered as a major factor limiting its predictability, a critical need for disaster mitigation associated with the trademark climatic swings of the ENSO. Improving climate models for ENSO forecasts relies on deeper understanding of the ENSO diversity but currently at a nascent stage. Here, we show that the ENSO diversity thought previously as 'complex,' arises largely as varied contributions from three leading modes of the ENSO to a given event. The ENSO 'slow manifold' can be fully described by three leading predictable modes, a quasi-quadrennial mode (QQD), a quasi-biennial (QB) mode and a decadal modulation of the quasi-biennial (DQB). The modal description of ENSO provides a framework for understanding the predictability of and global teleconnections with the ENSO. We further demonstrate it to be a useful framework for understanding biases of climate models in simulating and predicting the ENSO. Therefore, skillful prediction of all shades of ENSO depends critically on the coupled models' ability to simulate the three modes with fidelity, providing basis for optimism for future of ENSO forecasts.

Figure 1

Plot of combined EEOF of SST and SLP based on HADSLP data and ERSST data for the Pacific (30°S–30°N; 100°E–90°W) and with time lag of 18 months used to create the co-variance matrix (see text). (a) EEOF-1 pattern for SST at lag 0, (b) EEOF-1 pattern of SLP at lag 0, (c) Power spectra of principal component of mode 1:PC1; (d–f) same as (a–c) but for mode 2 (EEOF2/PC2); (g–i) same as (a–c) but for mode 3 (EEOF3/PC3).

Figure 2

(a) 31 year running correlation of PC1, PC2 and PC3 with June-September averaged (JJAS) rainfall over Indian region. (b) Lag correlation of PC1, PC2, and PC3 with July rainfall. (c) Composite of SST (shaded) and wind at 200hPa for the cases (months) when standardized PC1 is greater than +1 standard deviation (sd) during JJAS; (d) same as (c) but for PC2; (e) same as (c) but for PC3. (f), composite of SST (contours), rainfall (shaded) and wind at 1000hPa for the cases when standardized PC1 greater than +1 sd. (g,h) same as (f) but showing for cases when PC2 and PC3 greater than +1sd respectively. Signs are adjusted for the standardized PCs so that the values greater (less) than +1 (−1) indicates the El-Nino (La-Nina) case for the SST.

Figure 3

(a–c) Longitude-depth (2°S–2°N average) cross section of ocean temperature composite for the cases when PC1 > +1sd (standard deviation), PC2 > +1sd and PC3 > +1sd respectively. (d–f), similar to (a–c) for latitude-lag plot of the zonal mean 20°C isotherm depth for PCs > +1sd. The SODA reanalysis data (1871–2010) is used for this calculation.

Figure 4

(a) Growth of errors from peak El Nino to La Nina for Nino3.4, PC1, PC2, PC3; (b) Same as (a) but for peak La Nina to El Nino. Errors are defined as the standard deviation of the Niño index from all the sample days clustered at each lead time starting from the peak day.

Figure 5

(a) Evolution of the state of ENSO in the 3-D phase space of PC1, PC2 and PC3 between 1854 and 2004. (b–d) Same evolution but projected on 2-D phase spaces: (PC1, PC2), (PC2, PC3), (PC3, PC1).

Figure 6

Time-longitude section (10_S–10_N average) of different types of El-Niños and their reconstruction based on EEOFs. First two column shows super EL Niños (1972, 1982), the third column shows an EP El Nino (1965) and the fourth column shows a CP El Nino (1994) from Observations (a) and their reconstruction from EOF1, PC1) + (EEOF2, PC2) + (EEOF3, PC3) (b),(EEOF1, PC1) (c), (EEOF2, PC2) (d) and (EEOF3, PC3) (e) respectively. The reconstruction is based on EEOFs and PCs of these modes (refer text). Lag 0 is the peak month taken multivariate ENSO index data. Refer for the years selected for the study. The shading intervals are unevenly spaced.