stangandaho/Vultures-hotspot-areas-across-Africa

# Vulture Danger Zone Prediction Across Africa This project aims to identify and map vulture hotspot areas across Africa by analyzing risk zones using data from the African Wildlife Poisoning Database (AWPD) and a suite of natural and anthropogenic threat predictors. The analysis leverages the power of Google Earth Engine (code here) for geospatial data processing and R for statistical modeling and visualization. ## Data Source: African Wildlife Poisoning Database (AWPD) The core dataset for this study is the African Wildlife Poisoning Database (AWPD), which compiles records of wildlife poisoning incidents across the continent. The AWPD is managed by The Endangered Wildlife Trust and the Peregrine Fund. Access to the AWPD is restricted; researchers interested in using the data must request permission from the custodians. ## Project Overview The primary objective is to model and predict areas at high risk for vulture poisoning events. This is achieved by integrating AWPD incident data with a range of environmental and human-related predictors, such as land cover, proximity to infrastructure, and other spatial variables. The workflow involves data cleaning, exploratory analysis, feature engineering, spatial modeling, and the production of risk maps. Google Earth Engine is used to process large-scale geospatial datasets, while R is employed for statistical analysis, modeling, and visualization. ## Directory Structure The project is organized to facilitate reproducibility and clarity, with each directory serving a specific purpose in the analysis pipeline. The structure is as follows: - **datasets/**: This folder contains raw and processed datasets used throughout the project, including the AWPD data and any supplementary spatial layers required for modeling. - **scripts/**: All R scripts for data processing, analysis, and modeling are stored here. The workflow typically begins with `00_eda.R`. - **prediction/**: This directory holds the outputs of the predictiv …