FlexCast: Adaptive Weather Forecasting from Arbitrary Field Sets

2026-10-04

AI Summary

This study addresses the limited cross-configuration transferability of deep learning weather models caused by fixed variable sets. We propose FlexCast, an adaptive forecasting model that introduces a novel metadata-driven dynamic feature modulation and masked ensemble fusion mechanism. By integrating metadata-conditioned adapters, shared projection modulation, and a multi-scale U-Transformer architecture, FlexCast eliminates fixed-channel dependencies, enabling identity-aligned forecasting and recursive incremental prediction for arbitrary variable subsets using a single parameter set. Experiments on the ERA5 dataset demonstrate that the model exhibits superior cross-configuration generalization, while context compatibility significantly reduces forecast errors.

Abstract

Most deep learning weather models assign a fixed set of variables and pressure levels to predefined channels, limiting transfer across atmospheric field configurations. This dependence on a fixed field set limits the transferability of trained models across atmospheric field configurations. We propose FlexCast, a field-adaptive weather forecasting model that uses a single set of parameters to produce identity-aligned forecasts for variable-cardinality subsets drawn from a 69-field ERA5 registry. Specifically, a metadata-conditioned adapter the first encodes variable identity, pressure level, and field type and combines them with spatial features. Then, shared rank-16 projec?tions are modulated by metadata-dependent gates to produce field?specific features, while masked set fusion aggregates the available fields into a fixed-width representation. Subsequently, a multiscale U-Transformer processes the fused atmospheric features, while an identity-aware query decoder produces forecasts for the requested fields. Finally, FlexCast learns a standardized six-hour increment and applies it recursively to generate forecasts at longer lead times. Experiments on the 2020 ERA5 test set demonstrate that FlexCast operates across varying field configurations. Compatible cross-field context is associated with lower forecast errors, whereas mismatched context increases them.

Problem

• weather forecasting
• field adaptability
• transferability
• atmospheric configurations

Innovation

• field-adaptive forecasting
• metadata-conditioned adapter
• masked set fusion
• U-Transformer
• identity-aware query decoder