Aurora Forecasting

What is new in Aurora 1.5

Aurora 1.5 is a major extension of the Aurora Earth-system foundation model, developed by Microsoft Weather. It adds three headline capabilities: 22 additional forecast variables (spanning surface and pressure-level wind, temperature, humidity, precipitation, cloud cover, and radiation fields); hourly temporal resolution (up from 6-hourly); and probabilistic ensemble forecasting for uncertainty quantification. Together these make Aurora more useful for real-world decisions in energy, agriculture, transport, and climate-risk planning. Full details are in this technical paper.

Which new variables does Aurora 1.5 add?

The Aurora 1.5 update adds 22 new variables (21 new output variables and 1 new input variable) to Aurora’s original 4, including representative surface, pressure-level, wind, temperature, humidity, precipitation, and radiation fields. That broader coverage makes the model more relevant for sectors that depend on integrated Earth-system signals, from energy and agriculture to transport and resilience planning. Full details are in this technical paper.

What is ensemble forecasting, and why does it matter?

A single (“deterministic”) forecast gives one prediction of the future. Because forecasts are sensitive to initial conditions and model uncertainty, an ensemble runs multiple simulations to show the range and likelihood of possible outcomes. The ensemble version of Aurora 1.5 introduces stochastic perturbations to represent model uncertainty and generates multiple forecast members to estimate the spread of possible futures. For applications such as power systems, transport, agriculture, extreme-weather planning, and climate risk, the distribution of outcomes can matter as much as the single best estimate — enabling more confident, risk-aware decisions.

How is Aurora different from existing AI models like GraphCast?

Aurora differs from existing AI weather models like GraphCast, PanguWeather, and FourcastNet in a few key ways:

1. Generality: While models like GraphCast are designed for a specific task (10-day global weather forecasting at 0.25° resolution) using a single dataset (ERA5), Aurora is a more general system that can learn from many diverse datasets and adapt to various prediction tasks.
2. Scale of training data: Aurora was pre-trained on a much larger and more diverse set of weather and climate simulation data compared to models like GraphCast, allowing it to build more comprehensive general knowledge.
3. Architecture: Aurora employs a flexible encoder-decoder architecture with Perceiver modules that can handle datasets with varying resolutions, variables, and pressure levels, unlike task-specific architectures used in other AI weather models.
4. Performance: Aurora’s strong results across benchmarks demonstrate the advantages of its foundation model approach in terms of accuracy, computational efficiency, and ability to adapt to more granular resolutions and new tasks with less data.

What data was Aurora trained on?

Aurora was trained on a diverse set of weather and climate simulation data from various sources, including but not limited to:

1. ERA5: A high-quality global reanalysis dataset that combines model predictions with observational data.
2. CMIP6: Climate model simulations from the Coupled Model Intercomparison Project.
3. IFS forecasts: Predictions from the European Centre for Medium-Range Weather Forecasts’ Integrated Forecasting System at different resolutions.
4. GFS data: Analysis and forecast data from the National Oceanic and Atmospheric Administration’s Global Forecast System.

During pre-training, Aurora learned from over a million hours of this simulation data. For fine-tuning, Aurora used smaller, high-quality datasets specific to each prediction task, such as IFS-HRES data for weather forecasting and CAMS analysis data for air pollution prediction.

Which prediction tasks can Aurora currently tackle?

Currently, Aurora has demonstrated strong performance on several key atmospheric prediction tasks:

1. Medium-range global weather forecasting: Aurora can produce skillful 10-day global weather forecasts at both 0.25° and 0.1° resolution, outperforming the state-of-the-art IFS-HRES model and other AI models like GraphCast.
2. Global air pollution forecasting: Aurora can generate 5-day global forecasts of atmospheric chemistry and air pollutants at 0.4° resolution, matching or surpassing the accuracy of the CAMS operational system.
3. Extreme weather event prediction: Aurora has shown improved ability to predict extreme weather events like Storm Ciaran compared to other AI models, capturing sudden intensification that other models missed.

With the release of Aurora 1.5, several of these capabilities are now available: the model supports hourly forecasts, a broader set of 22 additional variables (covering wind, temperature, humidity, precipitation, cloud cover, and radiation), and probabilistic ensemble forecasting for uncertainty quantification. Aurora 1.5 has also demonstrated strong tropical-cyclone track forecasting. Ongoing work continues to expand Aurora toward a comprehensive Earth-system foundation model spanning atmosphere, ocean, and land.

How much computing power does Aurora use?

While Aurora can generate forecasts very efficiently once pre-trained and fine-tuned, the training process itself is computationally intensive. Pre-training Aurora on the diverse dataset of over a million hours of simulation data took about 2.5 weeks using 32 NVIDIA A100 GPUs. Fine-tuning is less demanding but still significant, taking around 5 days on 8 A100 GPUs.

However, this upfront computational investment pays off in the operational efficiency of the trained model. Aurora can produce a 10-day global weather forecast or a 5-day global air pollution forecast in just seconds on a single GPU, approximately 5,000 times faster than traditional numerical weather prediction systems like IFS, which require hours on large supercomputers.

Does this technology use Azure?

Yes, Aurora’s training pipeline has been optimized to leverage the cutting-edge capabilities of  Azure cloud computing for training deep learning models at scale. With Aurora 1.5, the model is also being made available through managed access via Microsoft Foundry and Microsoft Planetary Computer Pro for organizations that need additional data, infrastructure, and operational support. Aurora is intended to complement — not replace — physics-based models and domain expertise, and any downstream consequential use should include appropriate domain-specific validation.

Is Aurora open-sourced?

Yes. Aurora is available as an open research model under an MIT license. The code is open source on GitHub (https://github.com/microsoft/aurora (opens in new tab)) and model checkpoints are published on Hugging Face, so researchers and developers can evaluate, adapt, and build on the model.

Can I contribute to Aurora’s future development?

Our team is open to collaborate with domain experts to further enhance and expand Aurora’s capabilities, please get in touch!

What are the next steps for the Aurora project?

The Aurora team has several key next steps planned:

1. Open ecosystem: continuing to develop Aurora in the open (GitHub, Hugging Face) so the global community can evaluate and extend it.
2. Operational pathways: connecting the open research model to operational use through Microsoft Weather, Microsoft Foundry, and Planetary Computer Pro.
3. Collaboration with weather and climate agencies: working with partners such as the UK Met Office to explore how foundation models can work alongside established physics-based systems, from weather to climate time scales.
4. Enhancing capabilities: further improving resolution, accuracy, and uncertainty quantification, and expanding the range of prediction tasks.
5. Towards an Earth System foundation model: The success of Aurora in atmospheric modeling sets the stage for extending the foundation model approach to other Earth subsystems like oceans and land, moving closer to a comprehensive, unified model of the entire Earth System.