If the Met department says it’s going to be a bright sunny day, make sure you carry an umbrella. Or so went the jibe against the Indian Meteorology Department (IMD), often ridiculed for its wrong weather predictions.

Not any more.

The IMD’s nearly-accurate prediction about the cyclone Phailin, which hit Odisha on October 12, has not only raised its stock among the critics but has also placed it notches above foreign forecasters that predicted a super cyclone.

Often questioned for the credibility of its weather forecasts, IMD correctly predicted that Phailin would be a 'very severe cyclone' and refused to categorise it as a 'super cyclone'. It consistently maintained that wind speed of the cyclone would be 200 km per hour (kmph) against over-300 kmph predicted by many international agencies, including the US navy joint typhoon warning center and Britain's Met office.
The IMDF did not budge from its position even after a noted American meteorologist, Eric Holthaus, said the department was underestimating the potential winds and surges of Phailin.

Asked if the IMD felt vindicated over its forecast over the cyclone, the department’s director general, LS Rathore, said: “They (other forecasts) have been issuing over-warnings, we have been contradicting them. That is all that I want to say. As a scientist, we have our own opinion and we stuck to that. We told them that is what is required as a national weather service – to keep people informed with the reality without being influenced by over-warning.”

The credit for IMD’s turnaround, according to Rathore, went to a series of measures the department took in the past few years to get accurate predictions. “Since the 11th five year Plan, we have taken series of measures. We have been improving our observation system, which has helped to better define the initial condition. Development of human resources has also helped to a large extent,” he told reporters at a press conference after the cyclone had receded on October 13.

Under the 11th five-year plan (2007-2012), the central government had sanctioned a Rs 920-crore programme for modernisation of the IMD. Taking up the modernisation issue as one of its priorities, the government set up automatic weather stations, Doppler radars, connected them with most high-speed digital interconnecting systems and network as well as bought supercomputers for numerical weather prediction.

Doppler radars pick up evolving wind patterns and help forecasters issue short-duration predictions, called ‘nowcasts’ in weather parlance.
The government had also acquired four high-power computing systems (HPCS) to do the necessary number-crunching and run high-resolution global, regional and local numerical weather prediction models round-the-clock.

The need for IMD’s overhaul came after a string of inaccurate forecasts from the department in the last decade.  It failed to predict all the three droughts of 2002, 2004 and 2009.

According to a report in rediff.com, IMD went wrong not only in the first stage of the forecast that was issued in April 2009, but also in the "forecast update" issued in June, well after the onset of the monsoon. Its first forecast put the rainfall at 96 percent, the second one at 93 percent and the third one at 87 percent, whereas the actual rainfall turned out to be a mere 77 percent of the normal, resulting in drought over a large part of the country.

But the accurate Phailin cyclone prediction has charted a new direction for IMD.