As part of a multi-million-dollar deal, Spire Global is to provide a financial firm with its High-Resolution Weather Forecast model, which offers a six-day outlook powered by proprietary data collected from space, and develop an AI-powered model for long-range forecasting.
Training AI weather models
Spire will build on its recent collaboration with Nvidia to leverage Earth-2, Nvidia’s platform for weather and climate change modeling, to train and run AI weather models. These models are projected to increase the accuracy of weather forecasts and produce long-range probabilistic forecasts at unprecedented speeds.
Spire’s multipurpose constellation of satellites, using radio occultation technology, captures precise vertical profiles of temperature, pressure and humidity across the globe, including under-observed areas and remote regions. By using these proprietary data sets to initialize an AI forecasting model, the company aims to usher in a new era of accuracy, speed and reliability in weather forecasting.
Weather insights
“Having spent my career in the field of weather, I’m amazed at the rapid pace at which we’re seeing AI transform weather forecasting, especially when paired with proprietary data that can only be collected from space,” said Michael Eilts, general manager of weather and climate at Spire. “As climate change causes more volatile and extreme weather patterns, our models are equipping financial firms with the crucial insights needed to anticipate weather’s influence on price dynamics and market trends.”
In related news, the World Economic Forum recently released a report that reveals that Earth observation (EO) data has the potential to drive over US$3tn in cumulative economic benefits globally by 2030 while advancing a wide range of climate and nature solutions. Click here to read the full story.