Long-term stability and generalization of observationally-constrained stochastic data-driven models for geophysical turbulence. (10th January 2023)

• Record Type: Journal Article
• Authors: Chattopadhyay, Ashesh; Pathak, Jaideep; Nabizadeh, Ebrahim; Bhimji, Wahid; Hassanzadeh, Pedram
• Is Part Of: Environmental data science. Volume 2(2022)
• Publication Date: 2023-01-10
• Subjects: Data-driven climate model -- long-term stability -- transfer learning -- variational autoencoder; Environmental sciences -- Data processing -- Periodicals; 577.0285
• DOI: 10.1017/eds.2022.30
• Languages: English
• ISSNs: 2634-4602

Abstract

Abstract: Recent years have seen a surge in interest in building deep learning-based fully data-driven models for weather prediction. Such deep learning models, if trained on observations can mitigate certain biases in current state-of-the-art weather models, some of which stem from inaccurate representation of subgrid-scale processes. However, these data-driven models, being over-parameterized, require a lot of training data which may not be available from reanalysis (observational data) products. Moreover, an accurate, noise-free, initial condition to start forecasting with a data-driven weather model is not available in realistic scenarios. Finally, deterministic data-driven forecasting models suffer from issues with long-term stability and unphysical climate drift, which makes these data-driven models unsuitable for computing climate statistics. Given these challenges, previous studies have tried to pre-train deep learning-based weather forecasting models on a large amount of imperfect long-term climate model simulations and then re-train them on available observational data. In this article, we propose a convolutional variational autoencoder (VAE)-based stochastic data-driven model that is pre-trained on an imperfect climate model simulation from a two-layer quasi-geostrophic flow and re-trained, using transfer learning, on a small number of noisy observations from a perfect simulation. This re-trained model then performs stochastic forecasting with a noisy initial condition sampled from the perfect simulation. We show that our ensemble-based stochastic data-driven model outperforms a baseline deterministic encoder–decoder-based convolutional model in terms of short-term skills, while remaining stable for long-term climate simulations yielding accurate climatology.