IBM and its subsidiary The Weather Company have been named the ‘overall most accurate provider globally’ by ForecastWatch, a technology-driven weather forecast monitoring and assessment company.
Furthermore, the report data showed that the accuracy gap between The Weather Company and the next best overall provider has grown larger since the study began, from two times in 2017 to more than three times more likely in 2022, even with the addition of more weather providers studied.
Commissioned by IBM, the comprehensive ForecastWatch study spans six years, 2,182 global locations across eight regions, and 84 different accuracy metrics covering precipitation, temperature, cloud cover, wind and other forecast variables.
In its latest report, ForecastWatch analyzed around 250 million forecasts from 23 different weather providers. Each provider was noted each time it finished in first place for each metric and, according to the report, The Weather Channel recorded the most first-place finishes globally for every year studied and all forecast days of all the providers compared.
Sheri Bachstein, CEO of The Weather Company, said, “We’re seeing more companies jump into the business of weather forecasting as weather becomes more erratic and impactful. Despite increased competition, this report validates our continued innovation and leadership in weather accuracy. Our unique mix of ‘human and machine’ – advanced AI and technology plus human expertise – has helped us vastly outpace the field as we help people and customers make informed decisions based on insights from a recognized, trusted source.”
Artificial intelligence is a critical tool for meteorologists to make sense of billions of unstructured weather data points produced by satellites, ground sensors, radars, and more. With advanced machine learning algorithms and computing power from IBM, meteorologists at The Weather Company can use this data to simulate atmospheric conditions and better forecast the weather. Key technological developments include:
- A forecasting engine that uses AI to integrate input from nearly 100 forecast models around the world, weighing factors from each model based on geography, time, weather type and recent forecast accuracy, on demand.
- A significant breakthrough in global weather forecasting with its proprietary IBM GRAF (Global High-Resolution Atmospheric Forecasting) system, the first hourly-updating global weather model able to predict something as small as a thunderstorm.
- A unique collaboration of human and machine – proprietary technology called “human over the loop” forecasting that brings AI together with an expert team of nearly 100 meteorologists.
- Global radar leadership thanks to one of the world’s largest assimilations of weather radar data and technologies such as AI to use that data effectively.