Embedding -based Crop Type Classification in the Groundnut Basin of Senegal

2026-01-23

AI Summary

This study addresses the persistent challenge in smallholder farming regions of simultaneously achieving high performance, ecological plausibility, transferability, and accessibility in crop classification from remote sensing data. The authors propose a novel approach that leverages geospatial foundation model embeddings for crop mapping, systematically evaluated in Senegal’s Peanut Basin. For the first time, they demonstrate the comprehensive advantages of TESSERA embeddings in smallholder contexts. By integrating embeddings from TESSERA and AlphaEarth with time-series satellite observations, the method achieves a 28% improvement in accuracy over the next-best approach under cross-temporal transfer scenarios. It fully satisfies a four-dimensional evaluation framework encompassing performance, reasonableness, transferability, and accessibility, thereby substantially enhancing both the precision and practical utility of crop mapping in smallholder agricultural systems.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Machine Learning: Other Foundations of Machine Learning
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Crop type maps from satellite remote sensing are important tools for food security, local livelihood support and climate change mitigation in smallholder regions of the world, but most satellite-based methods are not well suited to smallholder conditions. To address this gap, we establish a four-part criteria for a useful embedding-based approach consisting of 1) performance, 2) plausibility, 3) transferability and 4) accessibility and evaluate geospatial foundation model (FM) embeddings -based approaches using TESSERA and AlphaEarth against current baseline methods for a region in the groundnut basin of Senegal. We find that the TESSERA -based approach to land cover and crop type mapping fulfills the selection criteria best, and in one temporal transfer example shows 28% higher accuracy compared to the next best method. These results indicate that TESSERA embeddings are an effective approach for crop type classification and mapping tasks in Senegal.

Problem

• crop type classification
• smallholder farming
• satellite remote sensing
• food security
• geospatial foundation models

Innovation

• embedding-based classification
• geospatial foundation models
• crop type mapping
• smallholder agriculture
• temporal transferability