As India progresses towards building Smart Cities, we at Smart Cities Council India spoke to Jatin Singh, CEO, of India's first private weather forecasting company – Skymet, about the tools required for good weather forecasting and action plan in case of any natural disaster.
What are the kind of technologies used in forecasting weather and likely success rate in predicting the actual outcome?
Weather forecasting is done by dynamical and statistical technique, in dynamical it's a computer model that works, while in statistical it is what happens climatologically. The forecast is based on short-term which is usually three days, medium-term (15-days) and long-term (entire season).
The accuracy of the event of rainfall (not the quantity) is about 70 per cent in case of short-term, 80-90 per cent in case of medium-term and again 70 per cent in case of long-term. Basically, monsoon is going to be right or wrong, if you do it for 10 years, seven would be right and three will go wrong.
Why is India lagging behind in prediction of weather as compared to western countries despite having a robust space program?
I disagree on this. I would say India is way ahead of the western countries when it comes to weather forecasting. We are very good at forecasting seasons. India behaves like a upper latitude country, you can find rainfall activity way in advance in terms of not just the timing but also the quantity. Precipitation in geographical areas is a combination of heat and water. You won't find any other tropical region which is better than India. Looking at the geography of the country this is the best we can do.
As India progresses towards developing Smart Cities across the country, what would be your advise to start up companies in technology on potential opportunities?
Engage with government and work on specifications - is the short answer. When we talk about technologies bulk of the business comes through tendering be it water, electricity etc. Amravati is one of the place to work for. Ideally, until the government gets its act together don't do anything.
Should urban planners consider developing a weather forecasting system for every city, and what would be the cost involved for the same?
Urban planners can do that. The cost depends on the city and kind of weather phenomenal you have battled. For in case of Delhi, you need less investment - because its relatively a dry place, Mumbai you need much more. For e.g. Jaipur and Delhi can share infrastructure which could be equivalent to Mumbai's requirement.
One thing that the country needs at the strategic productive level. At city level we need doppler weather radars, which gives you rainfall at every point. Every city should put together Rs 10-30 crore for that.
What would it take for our satellites to start delivering climate forecasts specific for cities/ zones/ regions?
Climate forecasting is over 20-30 years. Whereas satellites can only give you data and images. Satellites are doing what the can, you dont need more satellite work, but you need a lot of observation. At micro level we need to work on upper-end observations.
The satellites need to be equipped with GPS (Global Positioning System) - more specifically you need GPS-RO (Radio occultation) data to improve the weather forecast at a high resolution.
IMD has been proved right, probably for the first time with such accuracy. Why has it lagged behind?
The IMD's forecast this year was a success because it seems to have finally taken into consideration the strong El Nino correlation, which it ignored till last year.
What is your opinion on the use of drones in agriculture?
Its a good idea, its going very fast and also good for urban planning. We are already working on projects in Rajasthan, Gujarat and will soon work in Madhya Pradesh. The government should also think on putting proper aircrafts for the aerial photography because the problems with the drones is that they cover a very limited amount of area.
How do you relate this weather forecasting for upcoming Smart Cities projects, would it be better for predicting the disasters, so cities need to be disaster proof?
Cities are generally disaster resilient. In the last ten years I don't remember an earthquake destroying any city, but a flood has done it twice, once in Srinagar and another in Mumbai.
When it comes to floods, the kind of flood that causes havoc - at best can be forecasted only a day in advance and if you want a 90-100 per cent accuracy you would have a 4-5 hour window. And to do that it is necessary to have a working radar in these places. First is to forecast, second would be modelling the flow of water, from where and how will it flow - that is called Digital Elevation Model (DEM).
You should also have a good grid of actual rain-gauge data. First thing you need to do is put radars. The IMD has not been successful in running the radars network - there are 12 of which only six are working and the resolution is also very low. So the idea is that let the city build its own system and then integrate it to the national level.
In times of crisis, we have seen that there is a communication gap between the authorities and general public. How can you bridge this gap?
Firstly your scientific instrumentation radar needs to be working at all times. Secondly, there has to be a communication strategy to make it available on mobile, computer and television. Lastly, is the feedback strategy, you as a communicator of disaster should be able to give feedback to the authorities. That is where all the disaster management scenario should be worked at.
What opportunities do you see in Smart Cities as a weatherman?
We expect Smart Cities to allow private companies like us to take care of disasters, especially with the weather elements - both by putting in place the machines & communication strategy.
(Note - Image courtesy skymetweather.com)