mlsanityData to 7 Oct 2026

Hard-Constrained Deep Learning for Climate Downscaling

arXiv 2208.05424 ↗First posted 8 Aug 2022physics.ao-ph cs.LGcited by 29

Not on the map. Why?

Abstract (arXiv)

The availability of reliable, high-resolution climate and weather data is important to inform long-term decisions on climate adaptation and mitigation and to guide rapid responses to extreme events. Forecasting models are limited by computational costs and, therefore, often generate coarse-resolution predictions. Statistical downscaling, including super-resolution methods from deep learning, can provide an efficient method of upsampling low-resolution data. However, despite achieving visually compelling results in some cases, such models frequently violate conservation laws when predicting physical variables. In order to conserve physical quantities, here we introduce methods that guarantee statistical constraints are satisfied by a deep learning downscaling model, while also improving their performance according to traditional metrics. We compare different constraining approaches and demonstrate their applicability across different neural architectures as well as a variety of climate and weather data sets. Besides enabling faster and more accurate climate predictions through downscaling, we also show that our novel methodologies can improve super-resolution for satellite data and natural images data sets.

Key facts. Cited by 29 papers in the map (data through 7 Oct 2026). Builds on 19 of its 41 references that are in the map. First posted on arXiv 8 Aug 2022.

Builds on

19 of 41 references.

References · 19

  1. 2014Conditional Generative Adversarial Netscited by 2,152
  2. 2016Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Networkcited by 1,898
  3. 2017Enhanced Deep Residual Networks for Single Image Super-Resolutioncited by 1,543
  4. 2018ESRGAN: Enhanced Super-Resolution Generative Adversarial Networkscited by 1,250
  5. 2014Image Super-Resolution Using Deep Convolutional Networkscited by 1,218
  6. 2020Learning Texture Transformer Network for Image Super-Resolutioncited by 346
  7. 2017Video Frame Synthesis using Deep Voxel Flowcited by 209
  8. 2020SRFlow: Learning the Super-Resolution Space with Normalizing Flowcited by 209
  9. 2021DC3: A learning method for optimization with hard constraintscited by 181
  10. 2017DeepSD: Generating High Resolution Climate Change Projections through Single Image Super-Resolutioncited by 80
  11. 2020MeshfreeFlowNet: A Physics-Constrained Deep Continuous Space-Time Super-Resolution Frameworkcited by 78
  12. 2020Stochastic Super-Resolution for Downscaling Time-Evolving Atmospheric Fields with a Generative Adversarial Networkcited by 68
7 more
  1. 2019Achieving Conservation of Energy in Neural Network Emulators for Climate Modelingcited by 27
  2. 2020ClimAlign: Unsupervised statistical downscaling of climate variables via normalizing flowscited by 25
  3. 2021Fast and accurate learned multiresolution dynamical downscaling for precipitationcited by 18
  4. 2020A Fortran-Keras Deep Learning Bridge for Scientific Computingcited by 16
  5. 2021WiSoSuper: Benchmarking Super-Resolution Methods on Wind and Solar Datacited by 11
  6. 2020Augmented Convolutional LSTMs for Generation of High-Resolution Climate Change Projectionscited by 9
  7. 2022Physics-Informed Learning of Aerosol Microphysicscited by 9

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 1; 2024: 10; 2025: 13; 2026: 5.

29 papers.

Sort:

By area:

  1. 2024Fast, Scale-Adaptive, and Uncertainty-Aware Downscaling of Earth System Model Fields with Generative Machine Learningcited by 26
  2. 2024Generative Diffusion-based Downscaling for Climatecited by 25
  3. 2023Fourier Neural Operators for Arbitrary Resolution Climate Data Downscalingcited by 24
  4. 2024MambaDS: Near-Surface Meteorological Field Downscaling with Topography Constrained Selective State Space Modelingcited by 11
  5. 2024A Generative Framework for Probabilistic, Spatiotemporally Coherent Downscaling of Climate Simulationcited by 11
  6. 2025A Probabilistic Neuro-symbolic Layer for Algebraic Constraint Satisfactioncited by 11
  7. 2025EnScale: Temporally-consistent multivariate generative downscaling via proper scoring rulescited by 10
  8. 2024Evaluating the transferability potential of deep learning models for climate downscalingcited by 8
  9. 2024Local Off-Grid Weather Forecasting with Multi-Modal Earth Observation Datacited by 8
  10. 2024Projected Neural Differential Equations for Learning Constrained Dynamicscited by 5
  11. 2025RainShift: A Benchmark for Precipitation Downscaling Across Geographiescited by 5
  12. 2025DiffScale: Continuous Downscaling and Bias Correction of Subseasonal Wind Speed Forecasts using Diffusion Modelscited by 3
17 more
  1. 2025Efficient Kilometer-Scale Precipitation Downscaling with Conditional Wavelet Diffusioncited by 3
  2. 2025Assessing the Geographic Generalization and Physical Consistency of Generative Models for Climate Downscalingcited by 3
  3. 2025Generating realistic global precipitation fields from modelled atmospheric circulationcited by 2
  4. 2026Conformal Prediction for Generative Models via Adaptive Cluster-Based Density Estimationcited by 2
  5. 2026WIND: Weather Inverse Diffusion for Zero-Shot Atmospheric Modelingcited by 2
  6. 2026Quantizing Time-Series Models As Dynamical Systems: Trajectory-Based Quantization Sensitivity Scorecited by 2
  7. 2024Enforcing Equity in Neural Climate Emulatorscited by 1
  8. 2024Implicit Neural Representations for Simultaneous Reduction and Continuous Reconstruction of Multi-Altitude Climate Datacited by 1
  9. 2025On Global Applicability and Location Transferability of Generative Deep Learning Models for Precipitation Downscalingcited by 1
  10. 2025Field-Space Attention for Structure-Preserving Earth System Transformerscited by 1
  11. 2024A nudge to the truth: atom conservation as a hard constraint in models of atmospheric composition using an uncertainty-weighted correctioncited by 0
  12. 2025Vision Transformers for Multi-Variable Climate Downscaling: Emulating Regional Climate Models with a Shared Encoder and Multi-Decoder Architecturecited by 0
  13. 2025Physics-constrained generative machine learning-based high-resolution downscaling of Greenland's surface mass balance and surface temperaturecited by 0
  14. 2025A Flow-rate-conserving CNN-based Domain Decomposition Method for Blood Flow Simulationscited by 0
  15. 2025MeltwaterBench: Deep learning for spatiotemporal downscaling of surface meltwatercited by 0
  16. 2026A Differentiable Framework for Global Circulation Model Precipitation Bias Correctioncited by 0
  17. 2026Physics-Informed Super-Resolution of Atmospheric Datacited by 0