Project Grant 2615744
- The National Science Foundation Office of Integrative Activities awarded the University of California San Diego's Scripps Institution of Oceanography $516,390 on August 15, 2026, under the Geosciences program (CFDA 47.050) to develop mathematical and computational methods for calibrating atmospheric models used in weather forecasting and climate simulation. The project addresses the challenge of tuning dozens of adjustable parameters that control small-scale atmospheric processes such as cloud...
- The National Science Foundation Division of Mathematical Sciences awarded the University of Wisconsin–Madison $180,000 on July 15, 2026, under the Mathematical and Physical Sciences program (CFDA 47.049) to develop fast, flexible, and accurate covariance models for Gaussian processes in machine learning and artificial intelligence applications. The project addresses computational and methodological challenges in scaling Gaussian process models to large, high-resolution datasets by exploiting...
- The National Science Foundation Directorate for Mathematical and Physical Sciences awarded $250,000 to the University of Wisconsin – Madison on September 1, 2026, under the Mathematical and Physical Sciences program (CFDA 47.049) to develop high-fidelity generative mean flow models that advance scientific machine learning through a theory-validation feedback loop. The project addresses critical limitations in applying generative AI to physical simulations by building a new class of AI models...
- The National Science Foundation Division of Mathematical Sciences awarded the University of Wisconsin–Madison $250,000 on July 1, 2026, for statistical research on reliable estimation, prediction, and causal inference with AI-generated synthetic datasets under the Mathematical and Physical Sciences program (CFDA 47.049). The project develops statistical foundations and a unified theory for integrating multiple heterogeneous AI-generated synthetic datasets into scientific analysis while...
- This $399,162 Project Grant award from the National Science Foundation's Geosciences Program (CFDA 47.050) aims to develop interpretable, stable, and mass-conserving artificial intelligence (AI) models to improve the computational speed and efficiency of geoscientific models, such as those used for air pollution and climate research. The project will create simpler "surrogate" machine learning models for key components like atmospheric chemistry and wildfire plume rise, allowing for...
- The University of Wisconsin-Madison has been awarded a $336,926 project grant from the National Science Foundation Division of Atmospheric and Geospace Sciences under the Geosciences federal grant program (CFDA 47.050) to conduct research on the coupling mechanism of the Madden-Julian Oscillation and Quasi-Biennial Oscillation through cloud-radiative feedback. Over a three-year period from July 1, 2023 to June 30, 2026, the University will analyze satellite and reanalysis dataset observations,...
- This $185,163 Project Grant award from the National Science Foundation's (NSF) Mathematical and Physical Sciences Program (CFDA 47.049) supports the development of scalable Gaussian process methods for spatial statistics and machine learning. The project aims to create a universal toolbox for highly accurate and computationally efficient Gaussian process modeling to enable improved data analysis, prediction, and uncertainty quantification across diverse applications like carbon monitoring,...
- Federal Project Grant Award Summary The University of Washington received a $337,147 Project Grant award dated August 15, 2025, from the National Science Foundation's Division of Atmospheric and Geospace Sciences under the Geosciences program (CFDA 47.050) to conduct research on atmospheric predictability using machine learning (ML) frameworks. The primary deliverable is fundamental research that tests the conventional two-week limit of weather forecast skill by employing deep learning tools...
- This National Science Foundation (NSF) Project Grant award under the Geosciences Program (CFDA 47.050) provides $300,000 in funding to the Massachusetts Institute of Technology (MIT) from November 15, 2024 to October 31, 2027. The award supports the development of machine learning-powered "surrogate models" to increase the computational speed and efficiency of geophysical models used for air pollution and climate research. Key project objectives include: Creating simplified,...
- This $202,000 Project Grant was awarded on September 1, 2025, by the National Science Foundation's Division of Atmospheric and Geospace Sciences under the Geosciences program (CFDA 47.050). The award funds postdoctoral research through August 31, 2027, focused on bridging idealized atmospheric modeling with observational data to improve understanding of forest-atmosphere interactions over complex terrain. The primary deliverables include approximately 30 Large Eddy Simulation (LES) runs...
The National Science Foundation Office of Integrative Activities awarded the University of Wisconsin–Madison $377,437 on August 15, 2026, under the Geosciences program (CFDA 47.050) to develop mathematical and computational methods for calibrating atmospheric models using machine learning and automatic differentiation. The project leverages a differentiable atmospheric general circulation model to improve how atmospheric models are tuned—specifically, how dozens of adjustable settings controlling small-scale processes such as cloud formation are selected. Current calibration approaches are slow, expensive, and partly manual. By combining model gradients with machine learning, the work aims to increase model accuracy, stability, and interpretability, yielding methods applicable to other chaotic nonlinear systems including plasma physics, turbulence, and geophysics. The project will also train graduate students at the intersection of artificial intelligence and Earth science. The central hypothesis is that model gradients remain informative across long timescales relevant to atmospheric statistics. Research is organized around three questions: the spatiotemporal scales over which atmospheric gradients are informative, optimization approaches for noisy nonlinear hybrid models combining physical and machine-learned components, and optimal sub-grid representations. Performance runs from August 15, 2026, through July 31, 2029, at the University of Wisconsin–Madison in Madison, Wisconsin.
Mod # | Description | ReasonForModification | Federal Obligation | Date |
|---|---|---|---|---|
| Not listed | $377.4k | 8/14/26 |