VITA: Variational Pretraining of Transformers for Climate-Robust Crop Yield Forecasting
This work addresses the challenge of accurate crop yield forecasting for global food security, especially in data-scarce regions, by developing a domain-aware AI model that overcomes data limitations, though it is incremental as it builds on existing transformer and variational methods.
The paper tackled the problem of AI models underperforming in crop yield forecasting when yields deviate from historical trends by introducing VITA, a variational pretraining framework that learns from satellite-based weather data and transfers to limited ground-based measurements, achieving state-of-the-art performance with statistically significant improvements in predicting corn and soybean yields across 763 US counties, particularly during extreme years.
Accurate crop yield forecasting is essential for global food security. However, current AI models systematically underperform when yields deviate from historical trends. We attribute this to the lack of rich, physically grounded datasets directly linking atmospheric states to yields. To address this, we introduce VITA (Variational Inference Transformer for Asymmetric Data), a variational pretraining framework that learns representations from large satellite-based weather datasets and transfers to the ground-based limited measurements available for yield prediction. VITA is trained using detailed meteorological variables as proxy targets during pretraining and learns to predict latent atmospheric states under a seasonality-aware sinusoidal prior. This allows the model to be fine-tuned using limited weather statistics during deployment. Applied to 763 counties in the US Corn Belt, VITA achieves state-of-the-art performance in predicting corn and soybean yields across all evaluation scenarios, particularly during extreme years, with statistically significant improvements (paired t-test, p < 0.0001). Importantly, VITA outperforms prior frameworks like GNN-RNN without soil data, and larger foundational models (e.g., Chronos-Bolt) with less compute, making it practical for real-world use, especially in data-scarce regions. This work highlights how domain-aware AI design can overcome data limitations and support resilient agricultural forecasting in a changing climate.