Prof. K. Ashok is a professor at the Centre for Ocean, Earth and Atmospheric Sciences, University of Hyderabad. He is among India’s foremost monsoon scientists. His contribution to the field is foundational: he is a member of the research group that first discovered the Indian Ocean Dipole and identified the distinct climate pattern known as El Niño Modoki. He represented India on the UN team for the UNFCCC from 2010–2014 and served as an IPCC team member and contributing author for the organisation’s 2013 report.
While speaking to Rashmi Kumari of Neo Science Hub, Prof. Ashok explained the mechanics of the 2026 El Niño, why identical-looking El Niño years can produce opposite monsoon outcomes, the limits of seasonal forecasting, and the open questions he believes will define monsoon science over the next decade.
IMD lowered its monsoon forecast from 92% to 90% between April and May, and NOAA officially declared El Niño on June 11th. How does warming in the Pacific Ocean end up affecting a paddy field in Telangana?
Yes — IMD has been right on the mark yet again. Indeed, El Niños tend to cause deficit rainfall over the Indian summer monsoon region during summer. In meteorological terms, summer refers to June, July, August, and September — or simply June, July, and August, since this is the traditional scientific summer associated with the monsoon; during this period, we call it the summer monsoon.
When you ask about warming in the tropical Pacific, it’s important to note that this is not a typical or canonical El Niño. It appears to be developing into a very large event — its strength is already quite significant. This El Niño is characterized by anomalous warming in the eastern tropical Pacific and cooling in the western tropical Pacific.
Under normal conditions, it’s the opposite: the eastern Pacific is cool and the western Pacific is warmer. This is because of the easterly winds over the equator, which push water westward so that it piles up in the western tropical Pacific. These winds, combined with Earth’s rotation, also cause upwelling — cold, nutrient-rich water rises in the eastern Pacific, supporting a large fish catch.
That is the normal condition. But once in a while, an El Niño occurs, and both the ocean and atmosphere change accordingly. Normally, there is a lot of convection off the coast of Indonesia; during El Niño, this convection reduces. This anomalous pattern goes on to affect the Indian monsoon. First, there is less-than-normal rainfall: the overturning atmospheric cell known as the Walker circulation weakens and subsides, not only over the western equatorial Pacific but extending to the eastern equatorial Indian Ocean as well. This subsidence causes divergence, and in turn a convergence over the eastern equatorial Indian Ocean.
That convergence triggers divergence over the Indian summer monsoon region, producing anomalously dry conditions over India, depending on the strength of the El Niño. There are at least two or three additional mechanisms at play. One is the Rossby wave response: cooling over the Bay of Bengal, combined with the monsoon trough, produces an anticyclonic circulation. This is another well-known mechanism, and all of these processes operate together.
An El Niño typically starts around April or May and persists through the following winter, up to around February or March, after which most El Niños shift into a La Niña or neutral phase — this is the life cycle of an El Niño. Whenever an event is strong or moderate-to-strong, as in the current year — one of the strongest El Niños on record — it is no surprise that we are seeing dry conditions across India. Unless a strong counteracting force emerges, such as a positive Indian Ocean Dipole, these dry monsoon conditions are expected to persist through the summer monsoon season.
In 1997, a strong El Niño still brought a good monsoon. In 2015, a similar El Niño caused losses. This year, the Indian Ocean Dipole is neutral and expected to stay that way. Does that worry you more than the El Niño itself?
The 1997 case is actually what motivated our early research on the relationship between the IOD, the monsoon, and ENSO, which we documented in 2001. 1997 was also a very large El Niño. Looking at the modern record, 1982, 1997, 2015, and 2023 — along with the current year — were all major El Niño events.
In 1997, El Niño was widely expected to bring drought, since El Niño is one of the most forecastable climate phenomena, predictable at least a year in advance. By then, forecasting technology had advanced enough that people saw it coming, and with the memory of the severe 1982 drought that affected India, Australia, and other regions, people were bracing for a major drought or at least significantly deficient rainfall.
But, as you mentioned, what actually happened was a broadly normal monsoon by IMD’s definition. Some places received less-than-normal rainfall, but it was nowhere near as bad as expected.
Around that same period, in 1999, the Indian Ocean Dipole was discovered — by the research group I later joined. My own research, together with colleagues such as Professor Zhaoyong Guan and our supervisor Professor Toshio Yamagata, found that 1997 also coincided with a very strong positive Indian Ocean Dipole event.
A positive IOD is characterized by anomalous warming in the western Indian Ocean and anomalous cooling in the eastern Indian Ocean. It typically begins around April and peaks in November or December. In 1997, the IOD was strong enough to counteract the impact of El Niño. The anomalous cooling in the eastern Indian Ocean caused divergence there, which in turn produced convergence and rising motion over the head of the Bay of Bengal. Similarly, through a different mechanism, the warming in the western Indian Ocean caused anomalous convergence and upward motion over the Gujarat–Rajasthan region, extending into Pakistan.
As a result, along what we call the monsoon trough — running from the head of the Bay of Bengal through Rajasthan into Pakistan — strong positive IOD events, such as those in 1961, 1963, 1994, and to some extent 2019, tend to enhance rainfall along this belt. A negative IOD, by contrast, tends to suppress rainfall relative to normal.
So in 1997, the strong positive IOD reduced the impact of the co-occurring El Niño — a finding confirmed through our model experiments and later corroborated by other researchers using models and nonlinear dynamics. In 2015, however, there was no positive IOD or any comparable counteracting factor. That was the essential difference between 1997 and 2015: 2015 was a purely strong El Niño without any offsetting influence such as the IOD. People have since also discussed the Atlantic Niño and Niña as another possible factor, but that too was inactive in 2015.
This was the major factor behind near-neutral conditions in 1997 versus drought in 2015. One more interesting point: while El Niños tend to cause deficit rainfall during the summer monsoon, they have a tendency to enhance rainfall during the northeast monsoon. The 2015 Chennai floods, for instance, were partly attributable to the strong El Niño that year, along with other factors such as unusually warm conditions in the Bay of Bengal and urbanization.
This El Niño is expected to peak in September, right when the kharif crop is in its critical grain-filling stage. How serious is this timing, and what should we watch for in the August–September forecast?
Actually, El Niños typically peak not in September but around December — they begin peaking around that time. That said, this El Niño is already strong, and unless a positive Indian Ocean Dipole or dipole-like conditions develop in August or September, its impact will only intensify. July and August are the strongest monsoon months, but September carries its own importance, as you note. So El Niño will certainly have an effect, and it is not a good sign for the kharif crop.
Governments will need to manage their plans and strategies accordingly — and they have already taken note and put measures in place. Once September ends, the monsoon circulation shifts and the northeast monsoon begins to set in, at which point conditions will change. But at least through September and part of October, conditions are unlikely to be favourable for crops.
Monsoon was supposed to hit Kerala on May 26th but arrived on June 4th — the first missed onset forecast in 11 years, yet the seasonal forecast got the gross picture right. What does this gap tell us about the limits of monsoon science?
You need to understand the difference between weather forecasting and subseasonal-to-seasonal forecasting. Even back in 2009, when I was Chief of Climate Operations at the APEC Climate Centre, we ran more than twenty model outputs for global forecasts, and at that time there was essentially no skill in monsoon seasonal forecasting. We have come a long way since then, thanks to the efforts of the Ministry of Earth Sciences, and we are now getting fairly decent seasonal forecasts — even if, academically, they may not always meet strict standards.
The monsoon is a complex system that operates across many scales: daily variability such as monsoon depressions, intraseasonal variability such as the northward-moving active-break cycles, and even equatorial phenomena like the Madden–Julian Oscillation, which is strongest in winter but can still play a role in June, July, and August. Seasonal prediction skill largely comes from signals like El Niño, but that alone cannot tell the whole story — the onset of the monsoon and similar features depend on day-to-day weather and other factors, and tropical weather is inherently harder to forecast than mid-latitude weather, where large-scale dynamics dominate.
I would not call the Kerala onset forecast a weak forecast — a discrepancy of about a week is well within the nature of a seasonal forecast, as opposed to a day-to-day weather forecast, which IMD continues to update and refine. Planning agencies need to factor in this kind of week-scale uncertainty. One contributing factor this year was that El Niño was developing very rapidly. Models still have room to improve, aided by a growing wealth of observations, but factors like local weather and oceanic changes operate on timescales that may not be directly captured in a seasonal forecast. Given all of this, an onset forecast accurate to within about a week is something I would be quite satisfied with.
Climate change is bringing both longer dry spells and sudden intense rain. Is the monsoon changing? Is a “90% year” today the same as when you were a student?
Good question. A great deal of research has gone into understanding the philosophy of climate change, including the foundational work of Professor Syukuro Manabe, the Nobel laureate. As far as rainfall is concerned, a simple way to put it is: the rich get richer, and the poor get poorer. In other words, if you already receive more rainfall, you tend to continue receiving more; if you’re in a region with less rainfall, it tends to get even less. This isn’t universally true everywhere, but for monsoon regions and tropical areas with already substantial rainfall, it largely holds.
The underlying reason is a concept called saturation vapour pressure, which many students learn about even in classes 11 and 12, or in their early undergraduate years: as temperature rises, the atmosphere’s capacity to hold water vapour before it condenses increases. A rough rule of thumb is that for every one degree of temperature increase, extreme rainfall can increase by about seven percent — though this isn’t always exactly applicable. Higher temperatures allow more water to evaporate, and the atmosphere can store more moisture before reaching saturation and precipitating. This is why we tend to see more dry days — but once precipitation does occur, it carries more moisture than it would have in earlier periods, leading to stronger extreme rainfall events interspersed with longer dry spells.
This is, of course, a simplified picture; other factors play a smaller, sometimes offsetting, role. But overall, this is the broad picture associated with greenhouse-gas-driven global warming. Our own research, along with several other studies, has also shown that urbanization plays an additional role in intensifying these extremes.
So climate change is certainly increasing the intensity of rainfall extremes. As for whether a “90 percent year” today is the same as it was decades ago, that depends on the timescale you’re examining. If you look at extreme events specifically, their intensity has clearly been increasing. As far back as 2006, Goswami and colleagues published findings showing that normal and moderate rainfall events are decreasing in frequency while extreme rainfall events are increasing.
That trend continues today. But if you’re looking at total seasonal rainfall, the picture varies by region: in places like Chhattisgarh and Kerala, seasonal rainfall is decreasing, while in several other states it’s increasing. So it depends on where you are and on various regional factors — a topic for another discussion. What is clear is that rainfall patterns have changed seasonally, and so have the intensity and frequency of extreme rainfall events. And it isn’t only rainfall — droughts are increasing too, as are heat waves and moist heat waves.
Villages in Telangana are reporting unusually heavy jamun fruiting this year, traditionally seen as a sign of drought. Is there science behind this? Could a dry spring and a weak monsoon share a common climate cause?
This is an observation rooted in traditional farming knowledge. Take El Niño itself as an example — the name was given by fishermen based on their own observations, long before scientists studied the anomalous eastern Pacific warming. El Niño typically peaks around December, coinciding with Christmas, so fishermen, noticing the timing, named it after the Christ child. Similarly, farmers carry a wealth of observational knowledge that may not always be quantifiable in strictly scientific terms, but reflects genuine, accumulated experience.
As for whether a dry pre-monsoon period is necessarily followed by a good monsoon, the two aren’t always connected. Background warming from global warming has been a persistent factor in recent years, and a dry spring may well be linked to that broader trend. When El Niño is present, as it is this year, it can certainly contribute to dry conditions over Telangana, even though the correlation between El Niño and spring temperatures isn’t always strong. So I wouldn’t be surprised if this year’s dry spring is connected to the ongoing dry monsoon conditions.
Whether there’s a genuine connection between spring conditions and the monsoon is a good question, and the answer depends on which factors you’re examining. Take snow cover, for example: heavy snow cover over the Himalayas absorbs spring heat as it melts, which is thought to reduce the land–ocean temperature contrast that drives the monsoon. There are various theories along these lines, but for now, I think this year’s pattern may be somewhat connected to El Niño.
Hyderabad is home to IMD’s satellite data centre and hosted a recent crop-planning meeting. How well is the system moving from data centre to farmer decision this season, and where does it still fail?
I think, since IMD’s forecasts have improved considerably, thanks to a wealth of oceanic and land observations along with satellite improvements, dissemination has also improved. It isn’t just IMD’s regional centres providing information through bulletins; there are also apps and various other organizations contributing. Public awareness has grown too, including among the press, and information is percolating down to people far more than it used to.
New technologies are also emerging through mobile apps, and many independent agencies now share updates via social media. That said, there is still considerable room for improvement, particularly in tailoring forecasts to specific groups — not just farmers, but also urban residents and travellers, especially given the increase in extreme rainfall events and urban expansion. It’s common to see one part of Hyderabad receiving rain while another part stays dry. This is mostly a consequence of urban expansion rather than a fundamental change in monsoon structure, and it highlights an infrastructure challenge whenever extreme rainfall occurs — infrastructure is certainly one area we need to improve.
At the same time, forecasts need to be tailored to the local level — for instance, distinguishing whether it will rain in Gachibowli versus Jubilee Hills on a given evening. The technology for this is evolving: satellite data and radar systems are already in place, and with more radar coverage and better integration of local observations, both dynamical and AI-based forecasting can be improved further. AI has advanced significantly, and many of us are now working with it across different timescales. I expect substantial improvement in the coming years.
It’s worth noting that weather forecasting in the tropics is fundamentally harder than in the mid-latitudes. Here’s a simple analogy: imagine I give you a bag of ten thousand coins, and someone quietly removes 9,999 of them — you would notice immediately. That’s akin to the large-scale changes driven by mid-latitude dynamics, which are relatively easy to detect. But if I remove just a single coin from the bag, you likely wouldn’t notice. That’s the challenge of tropical weather prediction, where the Indian monsoon manifests in many different forms and scales.
Even so, we’ve come a long way. The devastating cyclones of 1977 and 1979 — including the Diviseema cyclone, which claimed many lives — motivated me to enter this field. Today, even as tropical cyclones continue to occur, casualties have become rare, reflecting real progress. I’m hopeful that similar improvements at the urban and local level will follow in the coming years.
What are the biggest unanswered questions in monsoon science? And which technology — AI, better models, or new satellites — will change forecasting the most over the next decade?
On the first question, there are many new research questions emerging. For instance, you might ask whether there’s a real connection between spring rainfall and the following monsoon. We may find a statistical relationship, but the deeper question is whether it’s genuinely physical, and whether it actually translates into better predictions.
Ultimately, the goal is to deliver forecasts that meet real-world demand — from industry, urban planning, and various other stakeholders — which means tailor-made forecasts requiring more and more observational data. Bridging the gap between traditional climate science and AI and machine learning is an area many of us, and others, are actively working on.
Another area I’m personally interested in is aerosols and emissions in urban areas. We understand their role in climate change reasonably well, but we haven’t fully explored their importance in weather prediction — some of us are now looking into that. There are also emerging questions tied to climate change, such as whether thunderstorm frequency and associated lightning-related fatalities are changing — these are examples of application-focused research questions.
Then there’s fundamental research: until 1999, we didn’t even know about the Indian Ocean Dipole — we discovered it. Now we’re exploring the role of the Atlantic Ocean as well. Understanding the basic science behind these phenomena remains essential, even as models and computing power continue to advance.
As for what’s needed to improve forecasting, better satellites and more observations are essential, but we should also focus on optimizing the use of the data we already have, along with local knowledge. There may be many untapped sources of data — for instance, mobile phone sensors that can measure temperature — and harnessing these kinds of innovations is an active area of work for many researchers.
Ultimately, as more people recognize the need to improve weather and climate forecasting, we need more people entering this field to expand its scope — in short, we need to build human capacity in this area.
