Now Reading
Why India’s Weather Department Couldn’t Predict the Chennai Rains Last Week

Why India’s Weather Department Couldn’t Predict the Chennai Rains Last Week

A waterlogged road in Chennai following the rain on December 30, 2021. Photo: PTI


  • Chennai received an immense amount of rainfall on December 30 – about 180 mm in a day. But the IMD had only predicted moderate rainfall.
  • Climate experts said the rain was likely due to a collision between an effect of the La Niña and an influx of moisture from the Bay of Bengal.
  • The area of interaction between these two entities was large but the effect was localised. This is one reason the IMD couldn’t have predicted the event.
  • This said, India’s weather prediction system also has some shortcomings; if they didn’t exist, the IMD may have been able to see the torrent coming.

Kochi: Unexpectedly heavy rains have been battering India’s east coast and parts of Tamil Nadu since December 30. The city of Chennai has borne the brunt so far: it received 180 mm of rain in less than half a day on the last day of 2021, bringing life in the city to a standstill. Roofs caved in, streets got waterlogged, and three people lost their lives in rain-related incidents.

Towards the end of the sudden spell, Tamil news channels were rife with frustration – by the people, commentators and some politicians – that the India Meteorological Department (IMD) had failed to do its job, to predict the rain or its intensity, putting the people in danger.

Climate scientists to whom The Wire Science spoke said the rains were possibly an anomalous event brought on by a combination of several weather changes. They also pointed out that the rain, while voluminous, lasted for only some hours, making its advent hard to predict, complicated by the fact that climate models have a limited predictive power.

La Niña

These rains were a “real freak event” that may not have been seen in a “long, long time” – possibly even in the past century – according to Raghu Murtugudde, a professor in the department of atmospheric and oceanic science at the University of Maryland, and currently a visiting professor at IIT Bombay.

“It is almost like a cloudburst – but cloudbursts do not happen in such flat areas,” he explained in an email. “They need orographic lift[footnote]I.e. lifted by a mountain[/footnote] to condense huge amounts of water very rapidly. The key ingredient to provide such large-scale dynamics is the La Niña.”

The La Niña is a climate phenomenon that is part of the normal, recurring patterns in the Pacific Ocean. It’s closely related to another pattern called the El Niño. In an El Niño year, the trade winds in the Pacific Ocean become weaker; in a La Niña year, the winds strengthen. Both patterns can disrupt large-scale air movements in the planet’s tropics by affecting the sea surface temperature and atmospheric pressure, and in turn affect the weather worldwide.

In a La Niña winter, Murtugudde said, there can be a drop in pressure from North India into the peninsula. A rush of moisture from the Bay of Bengal into Tamil Nadu probably ran into this pressure pattern. And the pattern acted like a mountain by mixing the cold air coming down with the warm moist air coming in from the warm bay.

“That’s a freak combination, which was made possible by the unusually strong winds from the east, coming all the way from the South China Sea,” he wrote.

The easterly winds – or the easterlies, one of the rain-bearing systems of the northeast monsoon that produce lots of rainfall over south India – have been around since November and are likely the result of the extended summer monsoon, which stretched into late October, according to Murtugudde.

“The strong vertical shear, or the change in winds from the surface to the upper atmosphere, have also suppressed the cyclones and have created an unusual rainfall pattern, with excess rain over peninsular India but drier conditions over the northeast.”

Indeed, the event was not a cyclonic depression but the result of an interaction of the existing easterly wave with strong westerly troughs, S. Balachandran, deputy director general of meteorology at the Regional Meteorological Centre, Chennai, told The Wire Science.

These pressure troughs are bands of low pressure that blow in from the west to the east. They’re also called westerlies, and often bring cool wind and rain.

According to him, it’s the season when easterlies are present in the lower level of the atmosphere.

“During the winter, the ocean will be warmer than the land. So when westerlies come in at the upper level at the same time, the interaction of all these factors causes such an event.”

Also read: The Complex Weather Phenomena Behind Tamil Nadu’s Intense Northeast Monsoon

No predictions of heavy rain

The IMD – which is the government department that warns of potentially damaging weather patterns, including heavy rain and cyclones – did predict the rain. But it only predicted moderate showers, not the torrent. Weather blogger Pradeep John, who also called the December 30 rains a “freak event”, tweeted that the sudden showers had caught them all by surprise.

“I want to apologise to you all, for missing to forewarn such an event to you,” he wrote. “This has never happened in the [last] 15 years and has caught us all by surprise. We have seen 5 mm forecast giving 50 mm rains, but not 200 mm rains.”

Tamil Nadu Chief Minister M.K. Stalin also said the meteorological department usually provided information on the likelihood of rainfall but that they could have failed to forecast the heavy showers. So what went wrong?

The heavy rains were a “sudden development”, Balachandran admitted. The area of interaction of the easterlies and westerlies was large, around 200-300 km wide. But the rains they produced were localised, affecting a relatively small area of the Chennai region.

According to him, limitations in predictive power keep the climate models that experts currently use from determining the precise location of impact – or explain why some events are localised in some areas and not others.

Predicting the weather

“The weather is a complex phenomenon where land is involved, as is the ocean and atmosphere, in very specific proportions. But climate models only give approximations,” Balachandran said. “That translation of heavy rainfall potential from a larger area to a localised one can be made only with the development of better science.”

India, like many other countries, uses global weather models that use computer simulations to predict weather in the country. These models produce forecasts for the whole world, for one or two weeks ahead. However, they trade scale for accuracy: they’re generally run at a lower resolution, which means both fewer forecast points per given area and at fewer time points.

However, Balachandran added, the inability to predict heavy rains is not a deficiency of the observation system.

“We have an integrated observation system, including satellites, automatic monitoring stations, radar, etc. to collect data. But these can give only approximate solutions until the science develops to understand these phenomena better.”

Then again, some of those advancements already exist.

The IMD model does indeed have a limited resolution, Murtugudde said. It is a global model, so resolution is limited to reduce the computational cost. If India’s model predicts rain at one point, that means rain within a 12 km radius; the model can’t ‘see’ variations inside that distance. It is similarly limited in the vertical space. But during the rains that Chennai received, the processes occurred at much smaller scales, which was probably why the IMD missed them, Murtugudde said.

“It also exploded quickly, within a few hours.”

Moreover, the way convection – the vertical movement of heat and moisture in the atmosphere – is represented in the models is probably not designed to capture small-scale processes, he added.

IMD director general Mrutyunjay Mohapatra told Hindustan Times that it’s difficult to attribute the rains to climate change because scientists will also need to study the atmospheric concentration of carbon dioxide, temperature profiles, etc. “But there is no doubt that extreme rainfall events due to climate change are increasing over India,” he added.

Chief Minister Stalin has said the Tamil Nadu government will take up the issue of upgrading weather forecast systems with the Centre. Meanwhile, the state is likely to have a rain-break January 3 until January 8 or so, John tweeted. And the IMD’s latest weather forecast, at 8:30 am this morning, ruled out severe weather over the city and predicted partly cloudy weather.

Note: This article was updated at 10:21 am on January 6, 2021, to correct a misplaced description of the westerlies.

Scroll To Top