Harnessing AI for Climate: IMB and NASA's Revolutionary Tool

Share
NASA and IBM create AI to help global warming (picture credit: NASA)
IBM and NASA have produced a highly adaptable AI tool aimed at tackling our planet's most pressing climate and weather prediction challenges

Amid escalating global weather unpredictability, climate change pressures mount — as does the need for sophisticated technological interventions to understand and predict atmospheric phenomena.

Artificial Intelligence (AI) has become an invaluable ally in this space, able to sift through colossal datasets, unveil patterns and forecast future climatic shifts.

This technological leap empowers scientists and decision-makers, permitting them to forge strategies that temper climate change's effects and pave the route towards sustainability.

A hallmark of innovation in this domain is the partnership between IBM and NASA in crafting a new AI foundation model.

This model represents a significant leap forward in climate science and technology, offering unmatched versatility and adaptability across various applications.

IBM and NASA: Pioneers in climate analysis

In collaboration with Oak Ridge National Laboratory, the AI foundation model is a highly- flexible tool adept at addressing an array of tasks from imminent weather predictions to extensive climate foresights.

Karen St. Germain, director of the Earth Science Division of NASA's Science Mission Directorate

Karen St. Germain, director of the Earth Science Division of NASA's Science Mission Directorate, says: "The NASA foundation model will help us produce a tool that people can use: weather, seasonal and climate projections to help inform decisions on how to prepare, respond and mitigate." 

Unlike many existing weather AI models, this model can be applied to various applications, including:
  • Creating targeted forecasts based on local observations: The model can generate more accurate and localised weather predictions by incorporating data from specific regions.
  • Detecting and predicting severe weather patterns: By analysing historical data and identifying patterns, the model can help identify and forecast extreme weather events such as hurricanes, tornadoes and heatwaves.
  • Improving the spatial resolution of global climate simulations: The model can enhance the accuracy of climate simulations by providing more detailed information at a finer spatial scale.
  • Enhancing the representation of physical processes in numerical weather and climate models: The model can help improve the accuracy of numerical models by better representing the complex physical processes that drive weather and climate patterns.

Unlocking the AI model's potential

The AI foundation model was pre-trained on 40 years of Earth observation from NASA's MERRA-2 dataset. This provides it with a well-versed knowledge of Earth's climatic dynamism.

Youtube Placeholder

Its design facilitates meticulous tuning, catering to both a universal and localised perspective.

Now accessible for download on Hugging Face, this tool diverges into two refined variants for distinct scientific and industrial applications:

Climate and weather data downscaling: This version can downscale climate and weather data to higher resolutions, enabling more localised forecasts and climate projections.
Gravity wave parameterisation: This version aims to improve the representation of gravity waves in numerical models, which can enhance the accuracy of weather and climate predictions.

The Impact of IBM and NASA's Teamwork

A collaboration dating back to the 1960s with IBM's computational contributions to NASA's Apollo missions, this enduring partnership has evolved and now leverages AI to mitigate climate change.

One key outcome of this partnership is the IBM watsonx.ai geospatial foundation model, built using NASA's satellite data.

It has also introduced an AI model trained on nearly 300,000 Earth science publications, facilitating organised, accessible scientific literature through IBM's PrimeQA.

Juan Bernabe-Moreno, Director of IBM Research Europe (UK and Ireland) and IBM's Accelerated Discovery Lead for Climate and Sustainability concludes: "This space has seen the emergence of large AI models that focus on a fixed dataset and single use case — primarily forecasting.

Juan Bernabe-Moreno, Director of IBM Research Europe (UK and Ireland) and IBM's Accelerated Discovery Lead for Climate and Sustainability

“We have designed our weather and climate foundation model to go beyond such limitations so that it can be tuned to a variety of inputs and uses.”


Receive the next edition of ClimateTech Magazine by signing up for its newsletter


As part of this portfolio, make sure you check out Sustainability Magazine and also sign up to our global conference series -  Sustainability LIVE


Also check out our Sister Brand, Energy Digital.

Share

Featured Articles

How Google, IBM & Met Office Use AI to Predict the Weather

AI is enhancing the accuracy and efficiency of weather forecasting and has the potential to mitigate the destructive impacts of natural disasters

What Does Vauxhall's Luton Closure Mean for UK EV Market?

The transition to EVs reveals conflicts between bold regulations and market challenges, prompting the industry to call for more flexible policies

How Henkel Turns Industrial CO₂ into Sustainable Adhesives

By creating sustainable water-based adhesives from captured CO₂ emissions, Henkel and Celanese are enhancing circularity in the packaging industry

How United Airlines is Revamping the Skies with SAF

Sustainability

Azerbaijan Invests in Carbonbit Technologies for DAC

Carbon Capture

Volvo Integrates 3D Tech for Electric Vehicle Innovation

Tech & AI