Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Ensembling geophysical models with Bayesian Neural Networks
Ushnish Sengupta, Matt Amos, Scott Hosking, Carl Edward Rasmussen, Matthew Juniper, Paul Young
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| BayNNE instead of multi model mean | ozone ensembling · Temporal extrapolation | RMSE (DU): 15.7 -> 4.4 | Improved |
| BayNNE instead of spatially weighted mean | ozone ensembling · South polar extrapolation | RMSE (DU): 19.6 -> 6.6 | Improved |
| BayNNE instead of spatially weighted mean | ozone ensembling · North polar extrapolation | RMSE (DU): 6.6 -> 4.7 | Improved |
| BayNNE instead of multi model mean | ozone ensembling · North polar extrapolation | RMSE (DU): 8.8 -> 4.7 | Improved |
| BayNNE instead of spatially weighted mean | ozone ensembling · Tropics interpolation | RMSE (DU): 5.5 -> 2.7 | Improved |
| BayNNE instead of spatiotemporal kriging | interpolation · Tropics interpolation | RMSE (DU): 7.0 -> 2.7 | Improved |
| BayNNE instead of bilinear interpolation | interpolation · Tropics interpolation | RMSE (DU): 31.2 -> 2.7 | Improved |
| BayNNE instead of bilinear interpolation | interpolation · Satellite voids interpolation | RMSE (DU): 1.7 -> 2.1 | Hurt |
| BayNNE instead of spatiotemporal kriging | interpolation · Satellite voids interpolation | RMSE (DU): 2.2 -> 2.1 | No clear effect |
| BayNNE instead of kriging | interpolation · Small features interpolation | RMSE (DU): 3.4 -> 3.2 | No clear effect |
| BayNNE instead of weighted mean | ozone ensembling · Temporal extrapolation | RMSE (DU): 8.7 -> 4.4 | Improved |
| BayNNE instead of multi model mean | ozone ensembling · South polar extrapolation | RMSE (DU): 30.5 -> 6.6 | Improved |
Closest papers by meaning
- Njord: A Probabilistic Graph Neural Network for Ensemble Ocean Forecasting (NeurIPS 2026 · similarity 0.84)
- Latent-EnSF: A Latent Ensemble Score Filter for High-Dimensional Data Assimilation with Sparse Observation Data (ICLR 2025 · similarity 0.84)
- Explanation of Dynamic Physical Field Predictions using WassersteinGrad: Application to Autoregressive Weather Forecasting (NeurIPS 2026 · similarity 0.84)
- Deep Random Features for Scalable Interpolation of Spatiotemporal Data (ICLR 2025 · similarity 0.83)
- Estimating Epistemic and Aleatoric Uncertainty with a Single Model (NeurIPS 2024 · similarity 0.83)
- Tyche: One Step Flow for Efficient Probabilistic Weather Forecasting (NeurIPS 2026 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
