‘Most severe weather can be predicted with reasonable accuracy’

Globally, weather causes close to 90 per cent of global disasters, impacting millions of people around the world every year and costing billions of dollars in damage. 2017 was one of the worst on record for catastrophic natural disasters, costing millions of lives and billions of dollars of damage across the world, says Himanshu Goyal , India sales and alliances leader, The Weather Company, an IBM Business.

Even the Indian subcontinent is highly vulnerable to natural disasters affecting millions of lives and impacting the economy. Combining accurate forecast data with AI, analytics, machine learning with smart sensors and Internet of Things (IoT) can go a long way in helping consumers and businesses make faster, smarter decisions based on upcoming weather.

These technologies have the potential to predict weather more accurately, so governments, communities and people can better anticipate and act on severe weather perils, says Goyal.

Excerpts:

Q: Is there a way technology can deal effectively with natural disasters?

A: The Weather Company, an IBM business, delivers personalised, actionable insights to consumers and businesses across the globe by combining the world’s most accurate weather data with industry-leading AI, IoT and analytics technologies.

It delivers billions of forecasts a day around the world via seamless cross-platform technology. The Weather Company connects newscasters, airline pilots, energy traders, insurance executives, state agency employees, retail management and more, to the weather intel they need, on any device. Additionally, we also aggregate the deepest, richest data sets - both business and consumer - to deliver personal, reliable and actionable weather information, analytics and insight.

This is a great testimony of how deciphering data using the right technology can really transform the way we can deal with natural disasters.

Q: What is the level of accuracy with which events can be predicted?

A: With the advances in weather observation, numerical weather models and an increase in the computing power, even the most severe weather can be predicted with reasonable accuracy. However, the depths of the insight are dependent on the scale of the phenomena. For example, we are more likely to forecast a tropical cyclone development at least three-four days in advance, in comparison to pinpointing the exact location of a tornado with the same lead time.

Q: Does India locally have enough data that can be mined to develop advanced weather forecasting models?

A: It is really hard to define what ‘enough’ is. From an observation standpoint, the more the variety, velocity and volume of data, the better it will be.

However, a very crucial aspect of data is also its quality and relevance which adds immense value. Companies need advanced capabilities and a strong technology platform to mine the data.

If you look at The Weather Company, our data story is about scale. We use more than 2.5 lakh personal weather stations, atmospheric data from 50,000 flights per day, more than 40 million pressure readings from mobile devices, and close to 162 forecast models.

All of this helps us determine and predict data with precision, accuracy and speed. There are several open-source numerical weather models which are available freely today.

However, their model configuration and data assimilation requires strong skills and experience, apart from a good observation network which is well calibrated and maintained.

Q: How big is the challenge to mine and crunch humongous data to make predictions?

A: There is a lot of data which is available on the Internet today. However, 80 per cent of the world’s data is not searchable. It lives behind corporate firewalls.

For organisations or individuals to mine this data into meaningful insights and to empower businesses, they need technologies such as AI or analytics to make sense of this data. So stand-alone raw data might not help, but one needs powerful tools to decipher it as well.

For instance, cloudbursts are usually caused by strong convective activities in a local scale, with short duration. Numerical weather prediction can capture the environment conditions conducive for convective development quite well, but to pinpoint the exact location and timing of the downpour is not easy sometimes.

While technology provides the right insights, it is important that multiple stakeholders come together to take prompt and timely action in the event of a natural calamity.

COMMENT NOW