AIFS-DOP: End-to-End Medium-Range Weather Prediction from Observations Alone with Machine Learning

2026/06/17 by Ewan Pinnington, Peter Lean, Mihai Alexe +8 · 1 voice
Physics and Astronomy · #physics.ao-ph

paper · pdf

Abstract

We introduce the Artificial Intelligence Forecasting System for Direct Observation Prediction (AIFS-DOP). AIFS-DOP is trained on a 40-year harmonized dataset of gridded observations, without using numerical weather prediction (NWP) reanalysis or model data. The resulting model is competitive with ECMWF's Integrated Forecasting System (IFS) when scored on a one year period of forecasts across 2021/2022. This progress on Direct Observation Prediction represents the first time that a data-driven model, trained solely on observations, is competitive with the IFS at medium ranges for several key upper-air and surface headline scores, when verified against observation data.

Citations

• Using data assimilation tools to dissect GraphDOP
• Learning Coupled Earth System Dynamics with GraphDOP
• Learning from nature: insights into GraphDOP's representations of the Earth System
• Huracan: A skillful end-to-end data-driven system for ensemble data assimilation and weather prediction
• AIFS-CRPS: Ensemble forecasting using a model trained with a loss function based on the Continuous Ranked Probability Score
• GraphDOP: Towards skilful data-driven medium-range weather forecasts learnt and initialised directly from observations
• Neural general circulation models optimized to predict satellite-based precipitation observations
• Probabilistic weather forecasting with machine learning
• Data driven weather forecasts trained and initialised directly from observations
• Learning skillful medium-range global weather forecasting
• Accurate medium-range global weather forecasting with 3D neural networks
• Deep Learning for Day Forecasts from Sparse Observations
• Forecasting Global Weather with Graph Neural Networks
• The ERA5 global reanalysis

Discussions

• Exciting new paper from ECMWF: ML weather forecasts directly from observations arxiv.org/abs/2606.19093 [bsky, 2 points, 1 comments]

Related

• AIFS -- ECMWF's data-driven forecasting system
• An update to ECMWF's machine-learned weather forecast model AIFS
• WP-MIP: An Artificial Intelligence, Hybrid, and Physically Based Model Intercomparison Project for Weather Prediction
• OMG-HD: A High-Resolution AI Weather Model for End-to-End Forecasts from Observations
• Data driven weather forecasts trained and initialised directly from observations