A Geospatial Approach to Predicting Desert Locust Breeding Grounds in Africa

TL;DR

This study develops a deep learning-based geospatial model to predict desert locust breeding grounds in Africa, leveraging multi-spectral satellite images and outperforming existing methods for early warning and control.

Contribution

It introduces a novel deep learning approach using multi-spectral satellite data and geospatial models for locust breeding ground prediction, achieving high accuracy without environmental data.

Findings

Prithvi-based model achieved highest accuracy (83.03%)

Multi-spectral images alone suffice for effective prediction

Models significantly outperform existing baselines

Abstract

Desert locust swarms present a major threat to agriculture and food security. Addressing this challenge, our study develops an operationally-ready model for predicting locust breeding grounds, which has the potential to enhance early warning systems and targeted control measures. We curated a dataset from the United Nations Food and Agriculture Organization's (UN-FAO) locust observation records and analyzed it using two types of spatio-temporal input features: remotely-sensed environmental and climate data as well as multi-spectral earth observation images. Our approach employed custom deep learning models (three-dimensional and LSTM-based recurrent convolutional networks), along with the geospatial foundational model Prithvi recently released by Jakubik et al., 2023. These models notably outperformed existing baselines, with the Prithvi-based model, fine-tuned on multi-spectral images from NASA's Harmonized Landsat and Sentinel-2 (HLS) dataset, achieving the highest accuracy, F1 and ROC-AUC scores (83.03%, 81.53% and 87.69%, respectively). A significant finding from our research is that multi-spectral earth observation images alone are sufficient for effective locust breeding ground prediction without the need to explicitly incorporate climatic or environmental features.