IBM forecasts Deep Thunder with machine learning & big data

IBM’s Weather Company has launched Deep Thunder, an advanced weather forecasting model which will use machine learning to precisely forecast weather impact on businesses. This model will also use historical weather data to improve weather forecasting.

Weather Company analyses about 100 terabytes of data on a daily basis which comes from different sources including third party sources, location data and through Weather Underground’s network of more than 195,000 personal weather stations.

The company claimed that it provides reliable weather forecasting by using enormous volume and variety of data with advances in atmospheric and computational sciences.

According to IBM, businesses around the world are now relying on accurate weather forecasting and Weather Company’s high performance forecast accuracy includes regional models that offer fresh guidance every three hours.

Deep Thunder models developed by IBM Research can be customised to the extent that business clients can pin-point weather condition for areas as small as 0.2 to 1.2 miles².

The forecasting model can take into account several factors such as vegetation and soil condition to better understand weather impact.

Deep Thunder model can also analyse weather data from retrospective points and use machine learning to create weather models that can predict how even subtle changes in weather can have significant impact on a business, consumer buying behaviour or how retailers should manage their logistics.

This is also very much relevant to insurance companies in analysing the impact of past weather events and calculate the validity of claims based on weather damage.

Apart from this, utility companies can also use this information to better understand the extent of damage caused to infrastructure including power lines and telephone lines during weather events and plan for the number of repair crews needed for the fix.

The Weather Company science & forecast operations head Mary Glackin said: "The Weather Company has relentlessly focused on mapping the atmosphere, while IBM Research has pioneered the development of techniques to capture very small scale features to boost accuracy at the hyper local level for critical decision making.

"The new combined forecasting model we are introducing today will provide an ideal platform to advance our signature services – understanding the impacts of weather and identifying recommended actions for all kinds of businesses and industry applications."