Indian Ocean Dipole
In September 1999, two independent research teams published papers in Nature describing the same previously unnamed phenomenon. What is Indian Ocean Dipole – and why did it take so long to notice? The answer to the first question: a coupled ocean–atmosphere pattern driven by contrasting sea surface temperature anomalies in the western and eastern tropical Indian Ocean. The Indian Ocean Dipole meaning for seasonal forecasting became clear quickly – it sits behind monsoon variability, drought and flood risk across three continents.
How the Indian Ocean Dipole Forms
The Indian Ocean Dipole (IOD) doesn't develop through a single trigger. It builds through a self-amplifying feedback loop between ocean temperature, surface winds and ocean circulation that, once started, reinforces itself until the seasonal cycle shuts it down.
Sea Surface Temperatures, Winds and Ocean Upwelling
Off the coast of Sumatra and Java, easterly winds drag surface water westward along the equator. When those winds strengthen, the process of upwelling intensifies – cold water from below rises to replace the displaced surface layer, cooling the eastern Indian Ocean. That cooling reduces atmospheric convection over the east, which shifts the wind pattern further, which drives more upwelling. The loop runs on itself.
One detail worth knowing about neutral conditions: in a typical year, the western Indian Ocean actually runs cooler than the east, because monsoon-driven upwelling in the Arabian Sea suppresses western temperatures. A positive IOD reverses this – the west warms while the east cools – which is the opposite of what the ocean's normal seasonal state would produce.
When IOD Events Develop, Peak and Decay
Most IOD events begin to take shape between May and June, when seasonal shifts in the monsoon winds create the initial ocean–atmosphere imbalance. By September through November, the pattern is usually at its strongest. December and January bring decay – the seasonal reversal of winds removes the feedback mechanism that sustained the anomaly.
Most IOD events decay by December. The 1997 positive event didn't – strong westerlies tied to an Intraseasonal Oscillation kept the ocean–atmosphere coupling active through what should have been the decay phase, extending it into February 1998. Across East Africa, that persistence meant floods. Across Indonesia and Australia, drought and fire conditions that the usual December shutdown would have ended instead ran on for months.
How to Read an Indian Ocean Dipole Diagram
Two rectangular boxes define the measurement zones on any standard Indian Ocean Dipole diagram – the western box at 50°E-70°E and the southeastern box at 90°E-110°E. The temperature contrast between them is what the diagram is built around.
Red and orange shading in the western box marks sea surface temperatures above the long-term average. Blue in the southeastern box indicates below-normal temperatures and a shallower thermocline – cold subsurface water sitting closer to the surface than usual. Arrows running along the equator show wind anomaly direction: during a positive IOD they point westward, reflecting the strengthened easterlies that drive upwelling off Sumatra.
Cloud and rainfall symbols cluster over the western box, where warm water pushes atmospheric convection upward. The eastern box shows suppressed cloud cover – the cool ocean below doesn't generate the same uplift.
Positive, Negative and Neutral IOD Phases
The three states of the Indian Ocean Dipole (IOD) produce broadly opposite climate outcomes across the Indian Ocean rim. The table below summarises the main differences.
| Phase | Western Indian Ocean | Eastern Indian Ocean | Wind anomaly | Main regional impacts |
|---|---|---|---|---|
| Positive | Warmer than normal | Cooler, shallower thermocline | Enhanced easterlies | East Africa: rain/floods; Indonesia/Australia: drought |
| Negative | Cooler than normal | Warmer, deeper thermocline | Anomalous westerlies | Indonesia/N. Australia: rain; East Africa: drought |
| Neutral | Near normal | Near normal | Near normal | Conditions close to climatology |
Positive Indian Ocean Dipole
Between June and November 2019, the positive Indian Ocean Dipole reached one of its highest recorded intensities. East Africa flooded – millions displaced across Kenya, Ethiopia and Somalia. On the opposite side of the ocean, Indonesia and Australia dried out. The 2019-20 Australian bushfire season burned more than 18 million hectares; the moisture suppression the IOD produced across southeastern Australia through spring was a primary driver.
Western Indian Ocean warming fuels convection and rainfall westward over East Africa. Off Sumatra and Java, enhanced upwelling cools the ocean surface, cutting off the moisture supply that Indonesian and Maritime Continent rainfall depends on. One structural feature of the IOD: positive events consistently run stronger in amplitude than negative ones – the ocean–atmosphere system amplifies the positive phase more readily than the reverse.
Negative Indian Ocean Dipole
In a negative Indian Ocean Dipole, the pattern reverses. The eastern Indian Ocean warms, the west cools, and anomalous westerly winds replace the strengthened easterlies of the positive phase. Indonesia and northern Australia receive above-normal rainfall; East Africa shifts toward drought conditions.
The 2021-22 negative IOD event was unusual in duration – a multi-year event that co-occurred with a multi-year La Niña, producing compounded effects on Australian rainfall and marine productivity. Most negative IOD events are shorter-lived and weaker in amplitude than their positive counterparts, which partly explains why research into negative phase impacts has lagged behind the positive phase literature.
Neutral IOD Conditions
Neutral is the default. For most of any given year, and across most years, the IOD sits in a state close to climatology – no meaningful SST contrast between the two boxes, near-normal winds, rainfall patterns driven by other systems rather than an active dipole.
A neutral IOD doesn't mean the Indian Ocean has no influence on regional weather. Other drivers – the Madden-Julian Oscillation, ENSO, local monsoon dynamics – operate continuously. What neutral IOD removes is the additional forcing that a strong positive or negative event would impose on top of those background conditions.
How the IOD Affects Rainfall and Weather
IOD effects depend on phase, intensity, timing relative to the monsoon season and how the signal interacts with other climate drivers. The same DMI value in June and in October can produce different regional outcomes.
The IOD and the Indian Monsoon
The 1997 Indian monsoon illustrates why the IOD–monsoon relationship resists simple rules. That year saw both a strong El Niño – which typically suppresses Indian monsoon rainfall – and a strong positive IOD. The two systems pulled in opposite directions, and the IOD's moistening effect on the western Indian Ocean partially counteracted the El Niño drought signal. The monsoon that year was below normal but not catastrophic, despite one of the strongest El Niño events on record.
The 2019 season ran differently. A strong positive IOD developed alongside a neutral ENSO – no El Niño counterforce to complicate the picture. Indian monsoon rainfall that year reached 110% of the long-period average, the first excess monsoon since 1994. The 1994 season also coincided with a strong positive IOD. The pattern is suggestive, but it isn't a guarantee: timing, monsoon onset dynamics and regional circulation all shape the final outcome.
Effects Across East Africa, Indonesia and Australia
The 1997 positive IOD brought floods to East Africa – thousands killed, hundreds of thousands displaced across Kenya, Ethiopia and Somalia, and a malaria outbreak that followed the standing water. The 2019 event landed in the same region with similar force: millions displaced, the floods arriving on the same seasonal schedule as before.
Indonesia sat at the cooler end of the same dipole. Eastern Indian Ocean cooling during both 1997 and 2019 cut off the moisture supply the archipelago depends on. Drought set in across Sumatra, Borneo and Java; fires followed, affecting tens of millions of people through two separate events separated by two decades but driven by the same mechanism.
Australia's response shifts with timing. A positive IOD peaking in September through November dries out the weather systems that would otherwise deliver spring rainfall to the southeast and southwest – the continent enters its fire season with less moisture in the landscape than a neutral year would leave behind.
Track seasonal weather forecasts and climate updates for your region on MeteoFlow.
How the Dipole Mode Index Tracks IOD Events
The Dipole Mode Index (DMI) is the standard numerical measure of IOD state – a single value that captures the east-west contrast the phenomenon produces.
What the DMI Measures
DMI is calculated as the difference between sea surface temperature anomalies in two defined boxes: the western equatorial Indian Ocean (50°E-70°E, 10°S-10°N) minus the southeastern equatorial Indian Ocean (90°E-110°E, 10°S-0°N). A positive DMI value indicates the western box is warmer than normal relative to the east – a positive IOD. A negative value indicates the reverse.
NOAA's Physical Sciences Laboratory publishes a monthly DMI time series at psl.noaa.gov/data/timeseries/month/DMI, updated regularly using the HadISST dataset. The index was developed by Saji et al. in 1999 alongside the original IOD identification papers and remains the reference measure used across seasonal forecast systems globally.
Why IOD Forecasts Do Not Guarantee a Specific Weather Outcome
A DMI reading above a threshold confirms that an IOD event is active. It doesn't confirm what rainfall will do at a specific location during a specific month. The 1997 season demonstrated this: a DMI well into positive territory coincided with an El Niño that partially cancelled the expected IOD impact on the Indian monsoon. The net outcome was neither what the IOD signal predicted nor what El Niño alone would have produced.
Timing is a second complication. An IOD that peaks in November, after the Indian monsoon has already withdrawn, carries far less influence on monsoon totals than one that develops by July. The DMI value at peak tells you about ocean state; the seasonal forecast models that translate that into regional impacts account for many additional variables the index itself doesn't capture.
How the IOD Interacts With El Niño and La Niña
El Niño years tend to produce positive IOD events more often than neutral years do – and La Niña years lean toward negative IOD. The statistical relationship is real. It isn't reliable enough to work in reverse: a positive IOD does not confirm an El Niño is occurring, and some of the strongest IOD events on record developed with ENSO sitting close to neutral.
When the two systems align, regional impacts compound. The 2021-22 multi-year negative IOD co-occurred with a multi-year La Niña, and Australia recorded exceptional rainfall and flooding across consecutive seasons – neither driver alone fully explains the duration or intensity of that wet period. When they oppose each other, as in 1997, the competing signals produce outcomes that neither system's textbook description would predict.
The IOD can also influence the Pacific through atmospheric teleconnections. Research has identified pathways by which a strong IOD alters the Walker Circulation in ways that feed back into ENSO development in subsequent seasons. The interaction runs in both directions: ENSO modifies the Indian Ocean background state that IOD events develop within, while the IOD can influence when and how Pacific anomalies evolve. Treating either system as purely independent of the other understates how coupled the tropical climate system actually is.
Follow seasonal climate indicators and regional weather forecasts on MeteoFlow.
FAQ
Can the Indian Ocean Dipole affect tropical cyclone activity?
Positive IOD events alter wind shear and moisture distribution across the Indian Ocean basin. Research links positive IOD phases to changes in cyclone track frequency and intensity in both the Arabian Sea and Bay of Bengal. The precise effect varies by region, season and whether ENSO is simultaneously active in the same direction.
Does the IOD change sea level across the Indian Ocean?
During a positive IOD, the shallower thermocline and enhanced upwelling in the eastern Indian Ocean lower sea level in that region, while warm water accumulation raises it in the west. Satellite altimetry has documented these shifts across multiple events. The pattern reverses during negative IOD phases.
Is "Indian Niño" another name for the Indian Ocean Dipole?
No. Indian Niño typically refers to basin-wide warming or cooling of the Indian Ocean as a whole – a uniform temperature shift rather than an east-west contrast. The IOD is specifically a dipole: the two poles move in opposite directions simultaneously. The terms describe different phenomena, though they are sometimes confused in non-specialist coverage.
Could climate change make extreme IOD events more common?
Research indicates that strong positive IOD events similar to 1997 and 2019 are expected to occur more frequently in a warmer climate, while moderate positive events may decrease. Negative IOD trends are less certain in current projections. The 2019 event's economic losses were estimated at US$558 billion globally – a figure that contextualises why frequency projections matter beyond meteorology.
