Kevin E. Trenberth and David P. Stepaniak
National Center for Atmospheric Research
Boulder, CO 80307 USA

To characterize the nature of El Niño-Southern Oscillation (ENSO), sea surface temperature (SST) anomalies in different regions of the Pacific have been used. We suggest that an optimal description of both the distinct character and the evolution of each El Niño or La Niña event requires at least two indices: (i) SST anomalies in the Niño 3.4 region (referred to as N3.4), and (ii) a new index we call the Trans-Niño Index (TNI), which is given by the difference in normalized anomalies of SST between Niño 1+2 and Niño 4 regions. A formal definition is given by

where the subscript "N" indicates normalized. The normalization factors in this case are the standard deviations of the anomalies over the climatological period 1950 to 1979 which amount to 0.92°C for Niño 1+2, 0.75°C for Niño 4, and 0.82 for the smoothed difference of the normalized anomalies. We use a five month running mean to smooth intraseasonal variability.

The N3.4 and TNI indices may be obtained at http://www.cgd.ucar.edu/cas/climind/TNI_N34/. They are also discussed more fully in Trenberth and Stepaniak (2001). The first index can be thought of as the mean SST throughout the equatorial Pacific east of the dateline, and the second index is the gradient in SST across the same region. Consequently they are approximately orthogonal. This relationship can be explored by computing cross correlations over about 20 year periods as a function of lead and lag up to ±20 months (see Fig. 1). In Fig. 1 the correlations are overall close to zero at zero lag. More revealing is the tendency of TNI to lead N3.4 by 3 to 12 months prior to the climate shift in 1976/77, and also to follow N3.4 but with opposite sign 3 to 12 months later. (A notable exception to this pattern is introduced by the prolonged 1939-42 El Niño.) However, after the 1976/77 shift, the sign of the TNI leads and lags are reversed, indicating that more recent El Niño events have first developed in the central Pacific and spread eastwards (e.g., Wang 1995). Prior to the 1976/77 shift, El Niño events tended to develop first along the coast of South America and then spread westwards (Rasmusson and Carpenter, 1982).

A sample of the type of SST patterns implied by these indices is shown in Fig. 2 which is a regression of global SST anomalies with TNI for 1900-1976. A negative boomerang shaped regression pattern of magnitude 0.2°C per unit standard deviation of TNI occupies the central tropical Pacific just east of the International Dateline, while a more localized positive regression of magnitude 0.8°C is centered on the Niño 1+2 region off the coast of Ecuador and Peru. Also present are regions of positive regression (up to +0.2°C) centered in the midlatitudes of the eastern Pacific in both hemispheres. For 1977-2000 (not shown) the centers of the two dominant structures occupying the tropical Pacific are shifted westward, and the sign of the patterns centered in the midlatitudes of the eastern Pacific are reversed. A secondary boomerang shaped region of positive regression also develops along the western edge of the Pacific rim during this period.

The TNI index is clearly involved in low frequency behavior of ENSO and the pattern is closely related to the so-called Pacific Decadal Oscillation. It has an advantage in that it is approximately orthogonal to N3.4 and applies on all time scales, and so does not require filtering of data to determine what is happening. Therefore, we suggest that it is essential to have at least two indices to describe the character and evolution of ENSO events, especially in studies which attempt to linearly remove the influence of ENSO using only a single index. Climate models have difficulty in realistically simulating ENS0 and a primary measure of success has been the magnitude of SST anomalies in the Niño 3.4 region. However, a realistic simulation of TNI also seems to be required to capture the different flavors of ENSO.