Never miss a great news story!
Get instant notifications from Economic Times
AllowNot now


You can switch off notifications anytime using browser settings.
The Economic Times
11,346.20-73.05
Stock Analysis, IPO, Mutual Funds, Bonds & More

Weather forecasting in times of extreme weather events

As extreme weather events become more common, demand for better forecasting is disrupting the entire space.

, ET Bureau|
Updated: Jun 16, 2019, 10.39 AM IST
0Comments
Agencies
weather-vane
A revolution in low-cost hardware, cheaper sensors and new technologies such as drones, combined with the growing power and effectiveness of algorithms are leading to new approaches in reading weather.
Professor Sridhar Balasubramanian has a personal weather station (PWS) installed on the terrace of his residence inside IIT Bombay’s leafy Powai campus. It measures rainfall, temperature, atmospheric pressure, wind speed, and wind direction.

Balasubramanian, an associate professor with the department of mechanical engineering at the elite institution, often climbs up an iron ladder to reach the equipment. The battery powered device sends its readings over WiFi to an electronic panel electronic panel, kept downstairs, in a room designated for his 11-month-old daughter, a room otherwise filled with toys and a swing.

Weather is more than a hobby for Balasubramanian. He teaches weather as a part of two mainline courses at the IIT. On June 12, as Cyclone Vayu flew past Mumbai towards the Gujarat coast over the Arabian Sea, Balasubramanian tweeted real-time predictions and satellite images of cloud formations.

The professor is part of a small network of weather enthusiasts in Mumbai, who have all installed a PWS. He is also member of a Whatsapp group called India Weather People, comprising of 100-plus enthusiasts. The data generated by the PWS at Powai is also shared with professional organisations such as the IBM’s Weather Company. Sometimes Maharashtra farmers reach out to him over Twitter seeking weather advice and timing of their crop sowing.

Personal weather stations, along with automatic weather stations, sensors mounted on ships, aircraft and other remote crafts and locations act as the nerve endings that feed information to weather prediction systems, which play an increasingly central role in our times, as extreme weather events appear to be becoming more common, thanks to climate change. The more number of weather stations that are part of a forecasting system, the greater the accuracy, by and large. The robustness of the prediction model is the other determinant of accuracy. In the US, for instance, the network known as Weather Underground is fed by a network of some 180,000 personal weather stations set up all over the country by enthusiasts.

climate-1

Balasubramanian has plans to formalise this network of Mumbai’s weather enthusiasts and start a new venture someday. When he does, it will join a thriving ecosystem in weather forecasting that is emerging in India, with a growing number of enthusiasts and their PWS’s, a large number of startups in the weather, satellite and agri-tech space, larger companies such as IBM’s Weather Company, Finnish firm Foreca, US-based Accuweather, and India-based Skymet and Weather Risk Management Services that are competing for a growing market for weatherrelated information and forecasting, and the state-run Indian Meteorological Department, whose long-standing pre-eminence in the space is increasingly under threat from nimble-footed enterprises in the private sector.

According Tracxn, a company that tracks deals data, there are more than 300 weather, satellite and agritech-startups in India that are either consumers of weather data or producers of weather-related information services.

The most important development in the space is that over the last few years, the monopoly of the government over weather data has ended, as private companies and networks have set up a large number of automatic weather stations around the country, made feasible by sharply declining costs of hardware and sensors. Some of these networks have been set up in collaboration with state governments. Satellite imagery by Indian Space Research Organisation (ISRO) is available through its commercial arm Antarix, as well as from other global sources. And the Indian Meteorological Department (a unit of the Ministry of Earth Sciences) has started opening up lately, sharing its own data freely.

All of this means that multiple weather forecasts are now available real time. It has also become possible to use the available data to provide independent, highly localised and granular forecasts that can be monetized. The new weather regime is way more democratic. “One no longer needs to wait for the 7.30pm Doordarshan News to know if it will rain tomorrow. We can now know if it will rain in an hour,” says Samuel John, chief operating officer of Satsure, a satellite imagery company that works with ISRO arm Antarix.

Big Weather
Satsure works as a partner for IBM’s Weather Company in India, offering satellite imagery combined with weather forecasts created by IBM. In 2018, it worked with state governments during the floods in Kerala and cyclone Titli that hit Andhra Pradesh, providing real-time updates and offering advice on the possibilities of fresh rains and affected areas that needed rescue or compensation.

The biggest disruptor in the space, however, has been Skymet Weather Services, started by Jatin Singh, a former broadcast journalist and son of a businessman who was a vendor to IMD. Singh started off in 2003 as a weather entrepreneur, as he saw an unmet need for weather data in news networks.

Skymet today has 6,500 automatic weather stations installed across the country, many of these in collaborations with state governments. It manufactures AWS, puts out weather forecasts for the entire country and sometimes challenges the IMD’s monsoon forecasts and methodologies, setting the two on a collision course that is being closely watched by stakeholders in the space. The criticism of IMD by various consumers of weather data and forecasts, including state governments, has been that it’s not been quick to change, or deliver granular and localized data relevant to specific regions.

climate-222

Some of them have started taking matters into their own hands.

Skymet has already helped the department of agriculture in Maharashtra to install 2,061 AWS across the state. Anil Salunkhe, deputy director at the state agriculture department, told ET Magazine the state government is setting up its own system of forecasts and advisories that can be issued to farmers, without depending on the IMD. Four agricultural universities in the state have been roped in to analyze the data churned out by the AWS’s. The project is ready to roll, and is awaiting the final go-ahead from the chief minister’s office.

Another state that has used Skymet’s expertise is Nagaland. State project officer Johnny R says that for about one and a half years, the Nagaland State Disaster Management Authority has been receiving data from Skymet to create forecasts and advisories on rain and storms for Nagaland that is disseminated through the Authority’s own YouTube channel. While this data is handed over to IMD, Johnny says that there is little data being generated by the IMD specific to the state.


IMD chief Mrutunjay Mohapatra, told ET Magazine in an interview over telephone from Geneva that collaboration, not competition, was the way forward. “It would be ideal if the private sector does not try to replicate our efforts as a rival but play complementary and collaborative role. We are the nodal body, but we cannot prescribe to others, neither can we discourage anyone. We have also made our data completely free and accessible on our website.”

Other start-ups also have their own data sets. The Weather Risk Management Services (WRMS), for instance, has set up more than 3,000 AWS’s across the country and data generated from these are shared exclusively with WRMS clients. WRMS is an IIT-Kanpur startup.

IBM’s Weather Company on the other hand works mostly with data generated by others. In a lean operation with only one meteorologist based in India, IBM uses IMD’s data plus other third party AWS’s and lets its international team analyse it. Both Skymet and IBM claim to have spotted the cyclone Fani that hit Odisha in May at least a fortnight before in made landfall. Cyclone spotting and advance warning is something that India has become good at. The IMD’s Mohapatra, is something of a celebrity in the weather forecasting circles for his talent in spotting cyclones. It has earned him the sobriquet ‘India’s cyclone man’.

Apart from the farm sector, information on storms and weather in general can be useful for different industries, especially logistics and airlines. Not to mention event organisers and suppliers around an event. For instance, the number of water bottles required at a cricket match can be predicted with some level of accuracy based on a forecast of temperature and humidity on the day of the match.

A revolution in low-cost hardware, cheaper sensors and new technologies such as drones, combined with the growing power and effectiveness of algorithms are leading to new approaches in reading weather.

IBM uses different sensors that are on board airplanes across the world to source data about temperature, pressure and wind direction. The Ratan Tata- and Softbankbacked, Boston-based startup ClimaCell Inc plans to use a combination of traditional forecasting, telecom signals, street cameras and drones to analyse and forecast weather. ClimaCell promises to make India one of its focus markets. Even barometers and motion or wind detectors on mobile phones are being roped in as sensors by some. Himanshu Goyal, who heads sales and alliances for IBM Watson Media and Weather in India says: “The challenge now is to make forecasts more granular, relevant to half a square kilometres of area,” he said.

Cracking the Monsoons
While India does well on cyclones tracking and there are enough indicators for predicting imminent rain even at the AWS level, the elephant in the room is of course monsoon forecasting. And despite a preponderance of weather stations and data, it remains a tricky affair. Anuj Kumbhat, the CEO of WRMS, says: “The topography of India is such that anything beyond a 7-day forecast is not very accurate.” Balasubramanian of IIT Bombay agrees. For 2019, while the IMD has predicted a normal monsoon, Skymet has gone ahead and predicted a less-thannormal rainfall.

Jatin Singh, Skymet’s founder, told ET Magazine that there are differences in the methodology followed by Skymet and IMD. He says: “If you track the development of the El Nino and the water temperatures of the Pacific Ocean, you can accurately predict monsoon rainfall in India on a piece of paper.” Not everyone agrees though, and Balasubramanian says one must also take into account the IOD or Indian Ocean Dipole, also known as the Indian Nino, a temperature variation of the eastern and western parts of the Indian Ocean.

IMD chief Mohapatra said the agency was working on improving its monsoon-related services. “Currently we provide district-level guidance for monsoon-related information across the country. We are also running a pilot in 200 blocks for block level guidance and will soon extend it to 6,600 blocks.”

Agritech, insurance and risk management bring in the moolah for these companies. Skymet gets 90% of its revenues from Agri-insurance. For WRMS, it is entirely weather risk management and preventive services. However, it is weather, monsoon and forecasting that seem to have captured public imagination. IBM, for instance, now runs a full-fledged course on weather at the Asian College of Journalism in Chennai. And at IIT-Bombay, professor Balasubramanian has come up with his own special two-day course on weather in July that is available for all and has already seen quite a few registrations. With personal weather stations now available at Rs 12,000 and upwards, and multiple Indian manufacturers for the same, weather and its forecasting too are becoming ‘close and personal’.
0Comments
Comments
Add Your Comments
Commenting feature is disabled in your country/region.
Download The Economic Times Business News App for Live Elections News & Results, Latest News in Business, Share Market & More.

Other useful Links


Follow us on


Download et app


Copyright © 2019 Bennett, Coleman & Co. Ltd. All rights reserved. For reprint rights: Times Syndication Service