Weatherbridge improves forecasts by adapting multiple weather data sources
WeatherBridge: Forecast-Aware Multi-Source Input Adaptation for Pretrained Weather Forecasting Models
Artificial Intelligence
Summary
Weather forecasting models often use a specific type of data called ERA5 to make predictions. However, in real-world use, they get data from other weather systems that don’t match ERA5 perfectly, which can hurt forecast accuracy. The authors created WeatherBridge, a tool that adjusts different weather data to look like ERA5 data without changing the forecasting model itself. This adjustment helps the model make better forecasts across various weather variables and time frames. WeatherBridge also works well with different pretrained forecasting models, making it flexible and useful for improving weather predictions.
What this means in practice
- •For weather forecasting teams: Adapt operational weather prediction inputs to improve accuracy without retraining the forecasting model.
- •For meteorological software developers: Integrate WeatherBridge to handle diverse initial state data sources and enhance model compatibility and forecasting quality.
Authors
Mutian Xi, Xinyu Su, Xiaohui Zhong, Hao Li
Abstract
Data-driven weather forecasting models are commonly trained on ERA5, yet operational deployment must use timely initial states from physics-based numerical weather prediction (NWP) systems whose distributions differ substantially from ERA5. Direct initialization with these states can degrade autoregressive forecasts, motivating input adaptation that keeps the pretrained forecasting model unchanged. Here, we introduce WeatherBridge, a unified, source-conditioned adapter that maps heterogeneous operational NWP initial states to ERA5 reference states without updating the pretrained forecasting model. WeatherBridge predicts the structured residual corrections with a Swin-UNet and injects source embeddings across multiple spatial scales, allowing the model to capture shared atmospheric structure and source-specific discrepancies. Its training objective combines latitude-weighted ERA5 reconstruction with One-step Rollout Alignment (ORA), which backpropagates one-step forecast error through a frozen FuXi model to optimize the adapted states for both initial state fidelity and downstream forecast compatibility. We evaluate WeatherBridge using daily GFS, HRES, and CMA initializations from 1 January to 22 November 2025 across six variables and lead times up to 240 hours. Relative to raw source initialization of FuXi, WeatherBridge reduces the mean RMSE of FuXi over the 6- 240-hour rollout for 17 of 18 source-variable combinations, with a median reduction of 10.2% and a maximum reduction of 46.3%. The adapter trained with frozen FuXi also improves FengWu forecasts on GFS inputs without retraining either the adapter or FengWu. These results demonstrate that forecast-aware input adaptation can improve forecasts from heterogeneous operational NWP inputs and transfer across pretrained ERA5-based forecasting models.