This work finds that a two-stage training curriculum, combining deterministic pretraining with probabilistic tuning, transfers well to downscaling, beating the state-of-the-art for RMSE.
Abstract
Climate data downscaling is the task of increasing the spatial resolution of climate data, typically by generating fine-resolution regional climate data from coarse global model output. Recent machine learning (ML) work in the related task of weather forecasting has seen significant improvements due to newly devised training methods and architectural components, but these have not yet benefited downscaling. We adapt two of these methods to create a family of lightweight probabilistic ML downscaling models built on a modified U-Net backbone and evaluate them on the CORDEX-ML-Bench suite for daily maximum temperature and precipitation across three geographic regions: the Alps, New Zealand and South Africa. We find that a two-stage training curriculum, combining deterministic pretraining with probabilistic tuning, transfers well to downscaling, beating the state-of-the-art for RMSE. Our work provides an advancement towards lightweight, probabilistic downscaling models, reducing the current trade-off between computational intensity and distributional fit.
A U‐Net‐based deep learning framework, the Joint Atmospheric fields Downscaling Network (JADNet), for rapid, joint downscaling of multiple atmospheric variables to kilometer resolution is introduced, offering a powerful tool for scalable high‐resolution weather and climate applications.
Hong-Xing Cui, H. Dasari, S. Sanikommu et al.· Journal of Geophysical Resea...· 0 citations
Effective climate risk assessment is hindered by the resolution gap between coarse global climate models and the fine-scale information needed for regional decision making. Existing downscaling methods struggle to bridge this gap: physics-based methods are too computationally expensive for large ensemble sampling of...
Zhong-Yi Wan, I. Lopez-Gomez, Rob Carver et al.· Nature Machine Intelligence· 0 citations
Results indicate that cross-attention conditioning offers advantages over simple concatenation for probabilistic precipitation downscaling, and that pre-trained foundation model representations may offer benefits in data-limited settings.
Victor Nascimento Ribeiro, Jorge Guevara, J. Moraga et al.· 0 citations
GCMagicc is a hybrid model that pairs a simple physical climate model with machine learning to generate ensembles of 10 climate variables at the resolution of full-scale Earth system models, without relying on GPU resources or retraining for new scenarios.
N. Meinshausen, M. Meinshausen, J. Lewis et al.· 1 citation
Extreme rainfall events are increasing in intensity and frequency as climate change accelerates. While kilometer-scale precipitation forecasts are critical for supporting local decision-making, the limited availability of high-resolution precipitation observations hinders their accuracy, especially in under-resourced r...
Pierre Lemaire, Luca Schmidt, Wietze Suijker et al.· 0 citations
The findings highlight the value of training strategies that allow models to directly learn bias correction during forecast transitions, emphasize the operational potential of combining sequential processing with near real-time discharge observations and identify physiographic catchment characteristics as key modulator...
O. Konold, Moritz Feigl, Patrick Podest et al.· Hydrology and Earth System S...· 3 citations
Exploring how generative AI could make machine vision more accessible to businesses. The post GenEye in a Box: Making Machine Vision Something You Can Just Ask For appeared first on GPT-Lab.
MIT News · Artificial Intelligence· news.mit.eduOct 7, 2026
Students in MIT’s Concourse program delve deeply into the human condition, debate challenging questions, and learn to develop judgment about issues that can’t be quantified.
Training AI agents with reinforcement learning can be challenging because their tools, context, and decision-making are managed by complex frameworks. Agent Lightning connects existing agents to RL training, making it easier to improve them without rebuilding them. The post Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses appeared first on Microsoft Research.
MIT News · Artificial Intelligence· news.mit.eduOct 6, 2026
