Predicting Crop Yield With Machine Learning: An Extensive Analysis Of Input Modalities And Models On a Field and sub-field Level

TL;DR

This paper presents a scalable machine learning framework for crop yield prediction that effectively integrates multiple input data sources, emphasizing the significance of modality selection tailored to specific regions and crops.

Contribution

It introduces a novel early fusion method for combining diverse input modalities in crop yield prediction at the sub-field level, applicable globally.

Findings

High-resolution crop yield maps enable precise training.

Sentinel-2 imagery combined with other modalities improves accuracy.

Optimal modality combinations vary by region and crop.

Abstract

We introduce a simple yet effective early fusion method for crop yield prediction that handles multiple input modalities with different temporal and spatial resolutions. We use high-resolution crop yield maps as ground truth data to train crop and machine learning model agnostic methods at the sub-field level. We use Sentinel-2 satellite imagery as the primary modality for input data with other complementary modalities, including weather, soil, and DEM data. The proposed method uses input modalities available with global coverage, making the framework globally scalable. We explicitly highlight the importance of input modalities for crop yield prediction and emphasize that the best-performing combination of input modalities depends on region, crop, and chosen model.