A Hybrid Modeling Framework for Crop Prediction Tasks via Dynamic Parameter Calibration and Multi-Task Learning

TL;DR

This paper introduces a hybrid modeling framework combining neural networks and biophysical models with multi-task learning to enhance crop prediction accuracy while maintaining biological realism.

Contribution

It presents a novel approach that parameterizes biophysical models with neural networks and employs multi-task learning for improved crop state predictions in data-limited scenarios.

Findings

Improves phenology prediction accuracy by 60%.

Enhances cold hardiness prediction accuracy by 40%.

Maintains biological realism in predictions.

Abstract

Accurate prediction of crop states (e.g., phenology stages and cold hardiness) is essential for timely farm management decisions such as irrigation, fertilization, and canopy management to optimize crop yield and quality. While traditional biophysical models can be used for season-long predictions, they lack the precision required for site-specific management. Deep learning methods are a compelling alternative, but can produce biologically unrealistic predictions and require large-scale data. We propose a \emph{hybrid modeling} approach that uses a neural network to parameterize a differentiable biophysical model and leverages multi-task learning for efficient data sharing across crop cultivars in data limited settings. By predicting the \emph{parameters} of the biophysical model, our approach improves the prediction accuracy while preserving biological realism. Empirical evaluation using real-world and synthetic datasets demonstrates that our method improves prediction accuracy by 60\% for phenology and 40\% for cold hardiness compared to deployed biophysical models.