Enhance Feature Selection for Spatiotemporal Modelling of blow flies Habitat using Google Earth Engine

Zaher, Eman; Moustafa, Marwa; Sobhey, Amira; Yones, Mona; Badawy, Rawda; Ma’moun, Shireen

doi:10.21608/eajbsa.2024.390168

Enhance Feature Selection for Spatiotemporal Modelling of blow flies Habitat using Google Earth Engine

Document Type : Original Article

Authors

¹ Department of Zoology, Faculty of Science, Zagazig University, Zagazig, Egypt.

² National Authority for remote sensing and space sciences, Cairo Egypt.

³ Entomology Department, Faculty of Science, Ain Shams University, Cairo Egypt.

10.21608/eajbsa.2024.390168

Abstract

Forensic Blow flies are of vital importance in the decomposition process. Their medical and veterinary importance came from their feeding habitat as they feed on faeces and cadavers. Also, they are vectors of viruses, bacteria and helminths. Recently, they were reported to cause myiasis in human and livestock. Google Earth Engine (GEE) is a cloud-based platform facilitating large-scale geospatial analysis through a web interface. It provides access to extensive satellite imagery and geospatial datasets, enabling computationally intensive studies. Feature selection serves as the cornerstone of successful habitat modelling. By carefully selecting relevant environmental variables, we can enhance predictive, reduce model complexity, and improve the understanding of species ecological requirements. Effective feature selection is crucial for addressing the challenge of high-dimensional datasets, especially in complex ecological modelling. While traditional methods have shown promise, identifying the optimal combination of feature selection techniques and predictive models remains an open research area. This study proposes a novel approach to model Blow flies’ distribution by leveraging the vast spatiotemporal data capabilities of Google Earth Engine (GEE) aimed to enhance habitat suitability accuracy and contribute to a deeper understanding of Blow flies ecology. Modelling Blow flies distribution will be extremely relevant in conserving wild and human life. Additionally, the represented data might serve forensic science.

Keywords