Abstract
The status of the El Niño–Southern Oscillation (ENSO), the primary source of seasonal climate predictability, is often represented by an index that measures sea surface temperature anomaly (SSTA) in the east‐central tropical Pacific, known as the Niño3.4 index. To better isolate ENSO timescale variations and physics, a relative Niño3.4 (rNiño3.4) index has been proposed as the Niño3.4 index departure from the global tropical mean SSTA. However, due to this subtraction, the rNiño3.4 index requires a potentially cumbersome variance adjustment to maintain consistent thresholds for defining ENSO events. Moreover, because the tropical mean SSTA is partially linked to ENSO, we could argue that the ENSO‐related component of the tropical mean SSTA should not be removed when defining the rNiño3.4 index. Here, we propose a “revised rNiño3.4” (rrNiño3.4) index, defined as the Niño3.4 index departure from the ENSO‐unrelated global tropical mean SSTA. An advantage of the rrNiño3.4 index is that it does not require variance adjustment, making it a physically rational alternative to the ENSO index. The connections to global precipitation and to equatorial Pacific recharge/discharge processes are generally comparable across the Niño3.4, rNiño3.4, and rrNiño3.4 indices. Furthermore, the cooling trends in the rNiño3.4 and rrNiño3.4 indices since 2010 are in line with the enlarged zonal SSTA gradients across the tropical Pacific and a La Niña‐like pattern. Thus, from this perspective, both the rNiño3.4 and rrNiño3.4 indices reflect the coupled anomalies between the tropical Pacific atmosphere and ocean on the seasonal‐to‐interannual and interdecadal timescales.