The Indian Ocean Dipole: The Climate Driver Closer to Home

El Niño receives the greater share of public attention, but there is a second oscillation operating considerably closer to the Indian coastline which exerts a comparable influence upon the monsoon, and which occasionally proves decisive.

The Indian Ocean Dipole, generally abbreviated to IOD, describes an oscillation in the difference in sea surface temperature between the western and south-eastern tropical Indian Ocean. It was identified relatively recently, in the late 1990s, and its recognition explained several monsoon seasons which had previously appeared anomalous.

The three phases

The IOD is measured by the Dipole Mode Index, which expresses the temperature difference between the western pole, in the Arabian Sea region, and the eastern pole, off Sumatra.

Positive IOD. The western Indian Ocean is warmer than normal while waters off Sumatra are cooler than normal. Enhanced convection develops over the warm western pole. This configuration tends to strengthen the moisture supply into the Indian monsoon and is generally associated with favourable rainfall over India.

Neutral IOD. The temperature gradient is near its long-term average and the dipole exerts little distinct influence.

Negative IOD. The western Indian Ocean is cooler than normal while waters off Sumatra are warmer. Convection is concentrated in the eastern pole, away from the Indian region, and rainfall over India tends to be suppressed.

Why the gradient matters

Convection develops preferentially over the warmest water. During a positive IOD, the warm anomaly in the western Indian Ocean is positioned within the pathway of the monsoon flow, close to the source region from which moisture is drawn towards India.

Enhanced convection there strengthens the moisture supply and reinforces the pressure gradient driving the monsoon circulation. During a negative event the arrangement is reversed, drawing convective activity away towards Indonesia and weakening the supply reaching the subcontinent.

Interaction with ENSO

This is the aspect which makes the IOD genuinely valuable for seasonal assessment, and it is the reason the two indices should always be considered together.

The IOD and ENSO are related but distinct. They frequently, though not invariably, occur in association: positive IOD events often coincide with El Niño, and negative events with La Niña. They can, however, occur independently, and it is these independent cases which prove most instructive.

Timing and development

IOD events follow a reasonably characteristic seasonal cycle. They typically begin to develop during May and June, strengthen through the boreal summer, reach peak amplitude around October, and decay rapidly during November and December as the seasonal reversal of monsoon winds disrupts the pattern.

This timing is significant for two reasons. The event develops as the monsoon becomes established, meaning its influence is exerted across the greater part of the season. And because it peaks in October, it also exerts considerable influence upon the north-east monsoon over Tamil Nadu and the south-eastern coast.

The rapid decay in early winter means IOD conditions offer limited predictive value beyond a single season.

Wider effects

The IOD influences more than Indian rainfall. Positive events are associated with drier conditions across Indonesia and Australia, and have been implicated in severe Australian bushfire seasons. They are also associated with increased rainfall over East Africa.

Within the Indian region, the IOD influences cyclone activity over the Arabian Sea, with warmer western Indian Ocean temperatures during positive events providing additional energy for tropical cyclone development.

Practical assessment

For anyone attempting to form a view of a coming season, I would suggest the following.

• Examine both indices together. Neither ENSO nor the IOD alone provides an adequate picture.
• Note that IOD development is observable from around May, which is close enough to onset to be informative for the season immediately ahead.
• Attend to magnitude, not only sign. A weak positive event carries far less implication than a strong one.
• Remember the limitations. Both indices describe seasonal-scale background conditions. Neither predicts the distribution of rainfall within the season, which is frequently what determines the outcome on the ground.

The IOD is monitored operationally by the India Meteorological Department, the Australian Bureau of Meteorology and several international centres, and current values are published regularly throughout the season.