Research map / Image generation / Super-resolution
Hard-Constrained Deep Learning for Climate Downscaling
Paula Harder, Alex Hernandez-Garcia, Venkatesh Ramesh, Qidong Yang, Prasanna Sattegeri, Daniela Szwarcman, Campbell Watson, David Rolnick
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 |
|---|---|---|---|
| add hard constraint layer (SmCL) to SR-CNN | SR-CNN · TCW4 | RMSE: 0.657 -> 0.582 | Improved |
| use soft constraining instead of no constraint | SR-CNN · TCW4 | RMSE: 0.657 -> 0.801 | Hurt |
| use soft constraining with larger alpha (ablation) | SR-CNN · TCW4 | RMSE: 0.237 (alpha 0.001) -> 2.618 (alpha 1) | Hurt |
| use multiplicative constraint layer (MultCL) instead of AddCL (ablation) | SR-CNN · TCW4 | RMSE: 0.580 -> 0.606 | Hurt |
| add hard constraint (AddCL) to GAN | SR-GAN · TCW4 | RMSE: 0.628 -> 0.602 | Improved |
| add MultCL to GAN (ablation) | SR-GAN · TCW4 | RMSE: 0.628 -> 0.732 | Hurt |
| add SmCL to ConvGRU (temporal) | SR-ConvGRU · TCW T1 | RMSE: 0.672 -> 0.500 | Improved |
| add ScAddCL to FlowConvGRU | FlowConvGRU · TCW T2 | RMSE: 0.673 -> 0.509 | Improved |
| add ScAddCL on out-of-distribution test data | SR-CNN · TCW OOD | RMSE: 0.661 -> 0.573 | Improved |
| add SmCL at 16x upsampling | SR-CNN · TCW16 | RMSE: 2.450 -> 2.368 | Improved |
| add SmCL on WRF with independent LR simulation | SR-CNN · WRF temperature | RMSE: 0.952 -> 0.950 | Improved |
| add soft constraint on WRF | SR-CNN · WRF temperature | RMSE: 0.952 -> 1.020 | Hurt |
Closest papers by meaning
- Quantile-Regression-Ensemble: A Deep Learning Algorithm for Downscaling Extreme Precipitation (AAAI 2024 · similarity 0.88)
- Precipitation Downscaling with Spatiotemporal Video Diffusion (NeurIPS 2024 · similarity 0.87)
- HR-Extreme: A High-Resolution Dataset for Extreme Weather Forecasting (ICLR 2025 · similarity 0.84)
- Satellite Observations Guided Diffusion Model for Accurate Meteorological States at Arbitrary Resolution (CVPR 2025 · similarity 0.84)
- DroughtSet: Understanding Drought Through Spatial-Temporal Learning (AAAI 2025 · similarity 0.83)
- Scaling Storm-Resolving Atmospheric AI Simulation to the Entire Planet (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.
