While the seasonal cycle of the Indian Ocean heat transport is relatively well studied, there are few papers focusing on its interannual variability. Furthermore, the studies that do exist consider variability over a relatively short time period. The Indian Ocean is associated with monsoons, which affect over one-third of the world population. It has been a debate for a long time whether monsoons play an active role in influencing the El Niño Southern Oscillations (ENSO), or if both are mutually interactive. The discovery of the Indian Ocean Zonal Mode (IOZM), the interannual aperiodic signal in the Indian Ocean, raises even more questions about the role of the Indian Ocean for both local and global climate variability. Recent studies suggest that the IOZM is closely related to the interannual variability of the Indian Ocean heat transport. Based on these considerations, we feel that understanding of the interannual heat transport variability in the Indian Ocean is essential for understanding the monsoon or even global climate variability.
For the current study, we use an intermediate 4½-layer thermodynamically active ocean model with mixed layer physics. The relative simplicity of the model makes it computationally efficient for studying interannual variability and allows both phenomenological studies and relatively easy interpretation of the physics. A comparison of the model’s mean seasonal cycle of the heat transport with previous studies confirms that the 4½-layer model reproduces other estimates very well. Furthermore, the high correlation of model IOZM variability with observational data demonstrates that the model produces reasonable interannual variability. Results of this study reveal that the Indian Ocean heat transport possesses strong variability at all time scales from intraseasonal (10–90 days) to interannual (more than one year). The seasonal cycle dominates the variability at all latitudes, and the amplitude of the intraseasonal variability is similar to the seasonal cycle. The amplitude of the interannual variability is about one-tenth of the seasonal cycle. Spectral analysis shows that a significant broadband biennial component in the interannual variability exists, with considerable coherence in sign across the equator. While the mean annual heat transport shows a strong maximum between 10°S and 20°S, interannual variability is relatively uniform over a broad latitudinal domain between 15°N and 10°S. The heat transport variability at all time scales is well explained by the Ekman heat transport, especially on the intraseasonal time scales. The inclusion of the Indonesian Throughflow does not significantly affect the heat transport variability in the northern part of the ocean. An increased southward (northward) heat transport occurs simultaneously with the growth of the IOZM positive (negative) mode.
The biennial heat transport variability seems to result from the internal Indian Ocean dynamics and varies from decade to decade. Wavelet spectra of heat transport show that the biennial signal is strongest in the heat transport in 1968-1988 and 1994-1998. Heat transport is shown to be strongly related to monsoon amplitude, as well as the IOZM and ENSO phase. The IOZM is shown to possess strong four-year variability between 1978-1988. During the time period considered, there is a statistically significant shift in correlations between heat transport and ENSO, as well as between heat transport and monsoon indices, and heat transport and the IOZM. Before 1980, the annually averaged cross-equatorial heat transport anomalies have strong positive correlations with All-India Rainfall Index (AIRI) and Webster-Yang Index (WYI), and strong negative correlations with ENSO and the IOZM. After 1980, correlations with AIRI, the IOZM, and ENSO decrease significantly, and the correlation with WYI becomes strongly negative. All these shifts are centered around the year 1980, the time when the correlation between ENSO and monsoon rainfall starts to decrease, which, according to several studies, is probably associated with the southeastward shift in the Walker circulation.
Based on the results obtained, we conclude that the internal variability of the Indian Ocean heat transport is strongly related to monsoon variability and the IOZM; however, at interannual time scales it is also affected by ENSO and decadal variability, such as the southeastward shift in the Walker circulation that occurred in 1980.