IMD works in real-time

IMD works in real-time


It the nodal agency of the Government of India responsible for providing weather and climate services for the safety of its citizens and also contribute for the economic and social development.

AVM (Dr) Ajit Tyagi
AVM (Dr) Ajit Tyagi
Director General of Meteorology

Can you tell us about the mandate of IMD?
It the nodal agency of the Government of India responsible for providing weather and climate services for the safety of its citizens and also contribute towards the economic and social development. It uses science and technology, and provides services to various sectors of the economy especially agriculture, aviation, power, health, defence, shipping etc.

What’s kind and scale of data provided by you to users?
We have end-to-end system, take observations across the country, that is, from Himalayas to the Port Blair and Andaman and Nicobar Islands, and collect the data in real-time. We take observations about various weather parameters, and have got a real-time communication system which receives this data. For example, we monitor daily rainfall and within 2-3 hours of recording, we are able to generate all India map of rainfall, that is, how much rainfall has taken place in different parts of the country. In case, there is a possibility of severe weather taking place like heavy rainfall, thunderstorm or strong winds, we are able to monitor it and also warn people against such adverse weather conditions.

We use supercomputers, state-of-art computers and various numerical models and then generate forecast in different time scales – short range forecasts, medium-range forecasts, long-range forecasts and seasonal forecast. Each forecast is then used by different sectors as per their application, that is, there will be different parameters and combinations in which a general forecast (rain, humidity etc.) may be used in sectors like agriculture, horticulture or aviation – so this is where the sector application comes. Then there is dissemination of information, whether we can disseminate information in real-time or not. This is most important as the warnings are short lived. So we have a real-time dissemination system. We use satellite based direct communication dissemination systems in coastal areas where warnings are given in local language to fishermen, that is, we generate audio warnings. So all these things are part of this total system, end-to-end site from observation to the final user. We use media, both TV and print for the purpose. For agriculture, we also use Krishi Vigyan Kendras (KVKs), the common service centers so that we can provide agro-advisories. The aggregate agro-services is anything which is related to weather and climate. Apart from this, we also have responsibility of providing warnings to Indian Ocean countries, and neighbouring countries about cyclones and storm surge. Because of climate variability and climate change, we are now trying to provide services which will be useful to manage the climate variability.

How are you making data easily accessible to users?
Up to recent times, everything was manual. The observatory used to take manual observations, which were then communicated through telegrams. These telegrams were then decoded and information was entered in registers, put in the form of charts, and reports were prepared which were then sent to people by post. This would mean a considerable amount of time. For example, the morning observation recorded at 8 o’clock could only be processed by evening, meaning 12 hours delay. Now everything is digital – most of the sensors we are making are automatic sensors, plus the transmission is through satellites or GPS. Thus, we receive information on time, and process it quickly with the help of supercomputers. We then analyse the charts, do the prediction, make the various products in the standard format and disseminate it at the same time via email, sms, voice etc. It is a process in real-time. In fact, it’s available on our website in real-time.

Agromet Advisory Servives (AAS) is much more than issuing just weather forecasts. It involves use of a wide range of tools for determining things like crop yield or soil mositure etc. Can you tell us more about it?
What we do is generate forecast. Forecast is starting point and the other end is the state of the crop, soil, etc. – these inputs are must and need to be synchronised. We generate all India forecast for all districts, that is, temperate, rainfall, humidity, wind, sunshine, clouds – all these things are made available by us for the next five days. This forecast is then sent to our state metrological centres which do the value addition depending on the local conditions and sends the new updated forecast to the agro metrological field units. The units consist of agronomists and soil scientists who have data about the crops of the region, area, soil, pests. With the help of the forecast, these experts advice people about irrigation, pest control or fertilisers that should be used by farmers depending on the composition of soil.

Union Minister Vilasrao Deshmukh recently said that climate change is a priority research area for the Ministry of Earth Sciences (MoES). What kind of contribution is IMD making in this direction?
The very basis of people talking about climate change comes from the data which we collect. When people talk about global warming, they are comparing it against something, that is, the data which was collected some 100 or 200 years ago etc. IMD is the repository of all this data, we have got data which is about 200 years old. So the entire talk of climate change starts with this data. For any modelling or things to run in real-time, you need climate data inputs, and that’s what IMD is going to provide. At present, we provide weather services, but we are soon going to start climate services as well. Climate services mean the forecast which is one year ahead, six months ahead, or three months ahead so that people can accordingly plan.

When are you going to start climate forecasting?
We all are working towards it. As of now, we have started providing some seasonal forecast, monthly forecast – they are a step towards this. As and when the skills for long range forecasting improve, we will begin the services.

Can you tell about the Indian Antarctic Programme?
It’s a programme which started some 30 years ago. IMD has been part of it right from first expedition. We take round the clock observations, round the year, and provide support to the base of forecast and also carry out various scientific studies. This is very important programme and has made our presence felt all over the world. We are now setting another station, Bharati – it should be operational by next year.

What is Background Air Pollution Monitoring Network (BAPMoN)?
It started in 1970s and ’80s and covers areas which are still not affected by pollution. The data collected from these areas are then used as a background to study how much changes are taking place in various atmospheric gases and pollutants like acid rain, or nitrous oxide etc.

Can you tell us about Satellite Distribution System (SADIS)?
SADIS has two components. One is having satellites for observing the weather. The other is using satellite for communications. Satellite dissemination is using satellite as a means to disseminate whatever data we collect, whether it is satellite data or other physical data we have to send to our centers in our country or neighbouring countries. That’s the satellite dissemination system.

You are the nodal agency responsible for monitoring seismic activity in and around the country? Can you tell us about National Seismological Network (NSN)? How does it operate?
It used to operate in a NLO mode, manual mode till recently where all our seismic stations spread all over the country, used to record data. There used to be a chart and a recorder which was very sensitive to the movements of the earth. The movement was recorded in the form of waves by the recorder. In case of earthquake, an expert would analyse the data collected by the recorder and determine its intensity. We would then correlate it with a number of neighbouring stations and try to determine the exact intensity and location of the quake.

Now we are working in real-time. As soon as the earthquake takes place, it gets recorded and is sent by the satellite communication. Similarly, information from other stations is collected at layer center processing system. This way, within five minutes, we can tell the epicenter of the quake. Also, we can give information about any earthquake which takes place over the oceans – the tsunami generated earthquake. Within 5-7 minutes of such earthquake, we are able to generate seismic messages to be sent to all disaster management authorities. Initial few hours are very important for any relief measure. We are able to provide such information round the clock.

All these activities involve day-to-day interactions between a number of agencies. How do you ensure smooth flow of information among them?
We work in a very collaborative mode. In fact, this is going to be the future. Right now, no department can do all the things of its own. For example, let’s take flood forecast. Flood forecasting in itself has got many components – rainfall, river flow and what’s going to be the rainfall in one or two days. While we can provide some information in this regard, it’s the Central Water Commission (CWC) which monitors it and finally gives the flood forecast. If we don’t provide them the details about the rainfall and the rainfall forecast, they cannot generate the forecast. So we work together. Earlier, we used to work separately resulting in the loss of critical time, but now we are working on sharing our data in real-time. That is, as soon as we take observation, the data is made available to CWC and whatever observations CWC takes, the data is sent to us. This way both of us benefit. We don’t want to set up observatory where they have observatory and they don’t have to set up data collection center where we have one. This is how we are working with various organisations, be it state governments, ISRO, various ministries like water resources, health, environment, etc. It has to be a collaborative effort so that national resources are shared.

Do you think NSDI will be great step in bringing various organisations together at one common platform?
As a format, yes it should be there. It’s a good platform to bring various government agencies together, where some protocols data can be shared.

Union Minister also talked about setting up National Monsoon Mission which aims to improve monsoon rainfall prediction in the country. Can you tell us about it?
National Monsoon Mission is a means to improve the forecasting system. At present, there is a big gap between the expectations of public and what we are able to provide. If somebody asks me to provide the forecast after a month in a particular state of India, I am not in a position to give that. I can provide the forecast for the whole of India but that makes no sense to a state chief minister or agricultural minister or anybody who is working in state. So what people are expecting us to do is to give them long range forecast or monsoon forecast of each state, down to the district. The forecast which they are expecting from us can only come with lot of improvements in our understanding of the monsoons and by using supercomputers, which we didn’t have till recently. So once we get these supercomputers, we are going to use the state-of-art models, hopefully in next 3-5 years, with lot of scientists working on these models. Then we will be in a position to provide forecast at the state level, the seasonal forecast or the monsoon. This is what monsoon mission is. Once such forecast is available, we will be in a better position to provide such services.

Why is predicting monsoons in India always a challenge?
Forecasting in itself is a challenge. Anywhere you go, be it US, Japan etc., forecast is always under criticism because as we improve our forecast, expectations increase.

However, the type of weather which we have in tropics is monsoon weather, which is highly variable in space, time and intensity. You may have a very heavy shower in one part of the city while the other part may be totally dry, the variability is immense. Suppose the forecast says there will be heavy rainfall in a particular state. The forecast will be right for the areas that receive the rain. However, in those parts of the city where there has been no rainfall, the forecast is wrong. So this is an in-built deficiency in appreciating the forecast. On the other hand, in Europe or America, weather is almost like a continuous factor and it follows a regular pattern and change. But in tropics, monsoon is sporadic and random, hence clouds can develop anywhere. It becomes difficult to precisely forecast the formation of clouds in a particular part of the city.

However, now that we have Doppler radars, we can monitor the cloud formation in a better way and can tell whether it is going to rain in south Delhi or north Delhi. The technology is going to come to our help in the short-range, very short range – 2 hours or 6 hours. Beyond that it will be the numerical model prediction which we are going to use for forecast. However, we cannot be 100 per cent accurate in tropics, so if somebody wants to know the exact, specific location or intensity of, say, rainfall, we may go wrong. But if we talk about weather forecast over an area, then yes, we have shown improvement. The accuracy has gone up. Earlier it used to be 60-65 per cent, now we are touching 75-80 per cent and we can easily go to 90 per cent in the next couple of years.

What about drought? Why do we go wrong in its prediction?
Drought is a very slowly developing phenomenon. It is not a cause of any overnight panic which the cyclone or heavy rainfall is capable of. Drought evolves slowly over a period of 3-5 weeks, so we can always give a direction, that is, an idea of stress that is building up and whether it’s going to evolve into a drought or not. Also, drought has got a lot many other implications. There are times when a state government may declare drought even when things are marginal, and there are times when it may not. For example, Punjab in 2009 and Haryana never declared drought even though the rainfall was less. So it depends on state governments as well.

Last year, IMD introduced an integrated forecasting and communication system for making accurate meteorological predictions? How is the progress on that front?
In the operational room, the forecaster is now able to see what is happening in any part of India and can communicate with those states. The situational awareness has improved tremendously. It is the backbone of our modernisation system where entire data gets integrated. With the observations, forecasts and tools available from various models and centers, we are easily able to generate forecast and communicate to the various agencies.

One of the reasons given in predictions going haywire is the use of outdated technology? How is IMD planning to overcome this problem?
In today’s world, where the competition is at global level, nobody would like to have second-hand technologies. And since we are paying, we certainly want to have the best. However, technology alone cannot help, it’s the use of this technology which is very important. In India, we have got highly capable people, so the question is how to use the technology to the best of our capabilities. If you go to African country or some other Asian country, they may not be able to use the same technology. Let me assure you that there is no second-hand or obsolete technology which we are using. We are using state-of-art technology which we will build on.

Now that the government has increased the budget for modernisation of IMD, what are your plans?
Budget is just one part of it, budget alone cannot help. It requires a lot of other things like the capacity building. Good quality people should come to the field of science. At present, they are opting for banking and IT sector. So if you don’t have good quality people, money cannot improve the forecast.

Modernisation is very important, and rightly needed. It is essential to have good quality observations as the scientific forecast is based on observations. Hence, good quality observations are required, only then good quality forecast can be expected. Modernisation has got the component of improving our observatory systems by means of Doppler radars, weather radars, automated stations which provide real-time data, and agro-digital stations. We use the inputs received by them in supercomputers to produce good models and generate better forecast. This is the total modernisation and of course, capacity building and human resource are an important part of it.

Earlier, there used to be perception that whatever we predict, it’s going to be opposite to that. But it is changing now, essentially because we have got tools now – there are radars, satellite and numerical models, so we are in a better position to forecast.

New technology involves training people as well. What’s being done in this regard?
It’s no longer enough to have knowledge only about science and not be aware about how the weather is going to affect various people. One needs to have an understanding of science, technology, communication and social needs of people. Basic foundation of science is must, but then one should know about the technologies which are being used – radars, satellites, remote sensing, computers and then be able to apply this knowledge into forecast. We have got people who are specialist in agricultural metrology, they generate information related to agriculture. Similarly, we have specialists in aviation sector, so we have people of specialised application source.

In what way is geospatial technology being used by you in improving your services to users?
The first and foremost is that it is a very user-friendly technology, so people are able to relate to it easily. GIS helps us put multiple data from multiple sources or handsets on a single map, which makes it easier for a decision maker to take decisions. For example, suppose we have data for rainfall, and we also have GIS map which shows the cyclone areas, the roads in those areas and the traffic flowing there. If we have got the real-time availability of the forecast of rainfall in that area, we will be able to determine the landslide prone areas. People would know the areas where there is an immediate danger. This way the technology makes it all very easy.

What kind of changes do you think we require in our system to be able to keep pace with today’s technological changes?
This is an evolving field. I think if we don’t evolve and change with what is happening outside, we will be a dinosaur. And any department, if it wants to be relevant to the society or to its own existence, it has to use technologies. But the only thing, the real challenge is the pace of change. It used to be very gradual and slow process till recently, but now it’s going to be a very fast. So we have to built-in the adaptability to accept these changes. Use the new technologies, albeit the benefit, use them as an opportunity rather than as an obstacle. We have to adapt and change. Earlier the change cycle used to be 15-20 days, now within 5 days, we have to overhaul our technologies and systems.