Heat Content Variability Related to ENSO Events in the Pacific

Abstract

Anomaly fields of upper-ocean heat content (OHC) in the Pacific basin from 30°S to 60°N over a 45-yr period are analyzed with special reference to their relationship with El Niño?Southern Oscillation (ENSO) events. In the present study, for convenience, the OHC is defined as vertically averaged temperature from the sea surface to 300-m depth. Based on a cluster analysis, it is found that interannual variability with a period of 3?6 yr is dominant in the eastern and western tropical Pacific, while multidecadal variability with a period of approximately 20 yr is dominant in mid- to high latitudes. In order to focus attention on the relationship between OHC anomalies and ENSO events, 3?6-yr variations are extracted by a bandpassed filtering with half-power points at 3- and 6-yr periods. The result clearly shows the cyclic and anticlockwise propagation of OHC anomalies on the circuit in the northern tropical Pacific. That is, an eastward propagation along the equator, a northward propagation at the eastern boundary, a westward propagation along the latitudinal belt centered around 16°N, and a southward propagation at the western boundary are found. The OHC anomaly spatially averaged over the entire equatorial Pacific from 4°N to 4°S leads the Niño-3 index by two seasons (i.e., six months), which is approximately one-quarter of the ENSO period or less. In addition, it is also found that the OHC anomaly in the entire equatorial Pacific is decreasing when the Niño-3 index is positive, while that in higher latitudes, from 5° to 20°N, in the tropical North Pacific is increasing at the same time. This fact indicates the exchange of an OHC anomaly between the equatorial Pacific and higher latitudes and supports the ?recharge oscillator? model for the ENSO dynamics. Further, it is shown that the magnitude of the OHC anomaly of the entire equatorial Pacific is directly related to that of the subsequent (two seasons or more later) Niño-3 index, which means that the larger amplitude of the OHC anomaly in the entire equatorial Pacific corresponds to the larger magnitude of the subsequent ENSO event, although an asymmetry exists between the preceding positive and negative OHC anomalies and the subsequent Niño-3 index. These facts strongly confirm the previous findings based on observations and numerical models. It is also shown that, after the 1976?77 regime shift, magnitudes of the OHC anomalies related to ENSO become greater and that an asymmetric feature between the preceding positive and negative OHC anomalies and the subsequent Niño-3 index is exaggerated.