Data-driven medium-range weather prediction with a Resnet pretrained on climate simulations: A new model for WeatherBench

Abstract

Numerical weather prediction has traditionally been based on physical models of the atmosphere. Recently, however, the rise of deep learning has created increased interest in purely data-driven medium-range weather forecasting with first studies exploring the feasibility of such an approach. To accelerate progress in this area, the WeatherBench benchmark challenge was defined. Here, we train a deep residual convolutional neural network (Resnet) to predict geopotential, temperature and precipitation at 5.625 degree resolution up to 5 days ahead. To avoid overfitting and improve forecast skill, we pretrain the model using historical climate model output before fine-tuning on reanalysis data. The resulting forecasts outperform previous submissions to WeatherBench and are comparable in skill to a physical baseline at similar resolution. We also analyze how the neural network creates its predictions and find that, with some exceptions, it is compatible with physical reasoning. Finally, we perform scaling experiments to estimate the potential skill of data-driven approaches at higher resolutions.

References in corpus (3)

• Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps
• Neural networks for post-processing ensemble weather forecasts
• Improving data-driven global weather prediction using deep convolutional neural networks on a cubed sphere

Cited by in corpus (16)

• Will Artificial Intelligence supersede Earth System and Climate Models?
• Sub-seasonal forecasting with a large ensemble of deep-learning weather prediction models
• SwinVRNN: A Data-Driven Ensemble Forecasting Model via Learned Distribution Perturbation
• Machine learning emulation of gravity wave drag in numerical weather forecasting
• Effective Training Strategies for Deep-learning-based Precipitation Nowcasting and Estimation
• Explaining the physics of transfer learning a data-driven subgrid-scale closure to a different turbulent flow
• Towards physically consistent data-driven weather forecasting: Integrating data assimilation with equivariance-preserving deep spatial transformers
• Equation-free surrogate modeling of geophysical flows at the intersection of machine learning and data assimilation
• A Practical Probabilistic Benchmark for AI Weather Models
• BAMCAFE: A Bayesian Machine Learning Advanced Forecast Ensemble Method for Complex Turbulent Systems with Partial Observations
• Interpretable multiscale Machine Learning-Based Parameterizations of Convection for ICON
• A Deep Learning Earth System Model for Efficient Simulation of the Observed Climate
• Explainable AI-Based Interface System for Weather Forecasting Model
• Global Climate Model Bias Correction Using Deep Learning
• Latent assimilation with implicit neural representations for unknown dynamics
• A comparison of data-driven approaches to build low-dimensional ocean models