'Urban meteorology is the next big thing'
Weather forecasting in India has come a long way from a shot in the dark to specific predictions with refined accuracies of the spot at which the event will happen. Whether it is predicting the scale and path of a cyclone even as the system is building up in the sea, or forecasting an approaching heat wave, the Indian Metreological Department (IMD) has redeeemed itself. Reently, the World Metreological Organisation patted IMD for its predictions on Cyclone Amphan.
The World Meteorological Organisation has once again lauded the IMD for its accurate prediction of the genesis, track and intensity of Cyclone Amphan, saying that lead of more than three days in forecasting helped authorities prepare, and thus mitigate the damage. This must be a great feeling
Yes, it is. It isn't the first time, though. We got our first big appreciation for predicting Cyclone Phailin which crossed the Orissa coast in October 2013. At that time, several western forecasts had gone wrong. The following year, we were appreciated for our forecasts for Cyclone Hudhud that struck near Visakhapatnam. Actually, our forecast was lauded in 2008, too, when Cyclone Nargis hit near the Yangong delta in Myanmar. Our forecast had a two day notice, but unfortunately, Myanmar was not able to act upon the warning in time, given its poor economic situation. So despite the warning, 1,40,000 people died, it is not a good memory.
How has IMD developed this expertise with cyclone warnings?
We began our modernisation programme in 2009, with a pardigm shift in approach from the analog to digital system. Earlier, we relied only on human observations, then we moved to automated rain guages. The charts which were plotted manually now are totally digitised and we get digital analysis of the data too. We installed Doppler radars and increased our computing powers. Today our supercomputers can manage 8.6 peta floating points operations per second. That is a huge change from the times when we had a single model which took upto 12 hours to process data into a forecast. Today, we can give a forecast within three hours. This was a Rs 435 crore project.
We've also deployed internet effectively to disseminate forecasts. We actually give a forecast with suggestions for actions to be taken – evacuate low lying areas, or kuchcha houses, keep away from the seas, for instance. Earlier, our forecast reached the district collector's office by telegram or telex after a lag of five hours, by which time, the cyclone would have advanced by 200 km. Now, the communication is instantaneous.
Everyday at 10:30 am, our 29 forecasters on cyclones meet via videoconference, discuss their observations and develop a consensus forecast.
How do you rate IMD's forecasting capabilities globally?
For cyclone forecasting, we are the best. We give our forecasts upto five days ahead, unlike ther forecasters. We also give specifics about how much rain or what winds to expect, when and where. We give suggested actions for every event. Monsoon and cylcones are our most important forecasts. Monsoon, because the nation depends so heavily on it. Cyclones, because they were once disastrous. Today, with good warnings, cyclones are only hazards in India, no longer disasters. The WMO has designated us as a regional specialised meteorological centre for tropical cyclones. We provide forecasting facilities to 13 countries along the Bay of Bengal and Arabian Sea, from Thailand through Pakistan to Yemen.
Has the Indian monsoon changed?
The monsoon is a large scale phenomenon and thus has large variations from year to year. There was a feeling that climate change was impacting monsoon, so we decided to study the data of the last hundred years. We found that there was no change in the overall rain between June to September. We next scrutinised the last 30 years' data in detail and noticed that within the monsoon season, there were changes. Some regions were getting less rain than before, namely the north east and Gangetic plains. Eastern Rajasthan, on the other hand, is getting more rain. We will need further studies to understand if this is because of climate change. We also saw that the number of days with intense rain was increasing, and the number of days with light rain decreasing. This is due to global warming. As the air warms, its capacity to hold moisture also increases. Rising temperatures also increase the rate of evaporation from the sea surface. So the convective clouds are dense and lead to very heavy rains.
You came with a new monsoon calendar this year, with changes in the dates of onset and withdrawal
That is a routine process, which was long overdue. The previous calendar was prepared in 1940, when we had just 140 rain gauge stations. Now, we have over 4,500 stations, we have better data. Also the climatology of today is different from the 1940s. Worldwide, the practise is to use latest data for deciding normals. Last year, we fixed the average amount of rainfall in India as 88 cm, based on data collected between 1961 and 2010. Soon, we will upgrade it with readings between 1971 and 2020. The monsoon calendar should also be updated in another ten years. If we continue with old dates, farmers will prepare for sowing at the wrong time. In a country like ours, so many events have to be synchronised with the weather.
But why do dates change?
That is climatology. Sometimes it is because we have better devices to get more accurate dates. Sometimes it could be because of climate change, but that we can only conclude after a proper study.
There are a lot of private forecasters now
Competition is always good. But they cannot replace us, we are the premiere organsiation, we collect the primary data. Most private operators use our data to come up with forecasts, and often, they are terribky off mark.
I believe that like the US, the private sector should focu on developing applications for agriculture and industry, insread of trying to replicate our work. They don't have the wherewithal for that. Weather forecasting is not a joke.
What is IMD's next big project?
We are working at providing impact based forecasts like we do for cyclones, for other weather systems, too. We have begun with heat and cold wave forecsts, but we want to refine that further. Also, give impact based forecasts for extremely heavy rainfall – such as which areas will be flooded, types of buildings at risk. This requires cooperation with other agencies to provide data on polulation, drainage, terrain and civic infrastructure.
Urban metreology is the next big thing. We have started with launching flood warning forecasts for Mumbai and Chennai this year. We also need to give gramin krishi mausam seva at the panchayat level, right now it is only at the district level. A country like India also needs a good multi hazard warning system. The country is so vast that we can be having cyclones and heat waves at the same time. Thus the need of a multi hazard forecast.