Efficient fine-tuning of 37-level GraphCast with the Canadian global deterministic analysis

TL;DR

This paper presents an efficient method for fine-tuning the GraphCast weather forecast model to replicate the Canadian GDPS system, achieving improved forecast skill with reduced training resources.

Contribution

It introduces a streamlined fine-tuning process that adapts GraphCast to a new analysis system using minimal additional training and computational effort.

Findings

Significant forecast skill improvement over unmodified GraphCast.

Effective adaptation achieved with only 37 GPU-days of training.

Model outperforms operational forecasts in the troposphere from 1 to 10 days.

Abstract

This work describes a process for efficiently fine-tuning the GraphCast data-driven forecast model to simulate another analysis system, here the Global Deterministic Prediction System (GDPS) of Environment and Climate Change Canada (ECCC). Using two years of training data (July 2019 -- December 2021) and 37 GPU-days of computation to tune the 37-level, quarter-degree version of GraphCast, the resulting model significantly outperforms both the unmodified GraphCast and operational forecast, showing significant forecast skill in the troposphere over lead times from 1 to 10 days. This fine-tuning is accomplished through abbreviating DeepMind's original training curriculum for GraphCast, relying on a shorter single-step forecast stage to accomplish the bulk of the adaptation work and consolidating the autoregressive stages into separate 12hr, 1d, 2d, and 3d stages with larger learning rates. Additionally, training over 3d forecasts is split into two sub-steps to conserve host memory while maintaining a strong correlation with training over the full period.