Exploring the trade-offs between spatial and spectral resolution in mapping invasive alien trees

Description This page contains the code and dataset used in the study "Exploring the trade-offs between spatial and spectral resolution in mapping invasive alien trees". The aim of this work was to compare different methods to improve the accuracy of mapping invasive alien trees. This specific research focused on evaluating the trade-offs between spectral and spatial resolution when classifying invasive alien trees using four different types of imagery. Many resource-constrained regions are heavily invaded by alien plants, particularly trees, and monitoring is crucial for their effective management. Yet there are often not funds to pay for monitoring. For these regions, it is important to understand trade-offs between spatial and spectral resolution to determine which freely available sensors provide the most accurate mapping results. This research specifically evaluated four types of imagery: 1. Aerial Photographs – 3 bands, 0.25 m spatial resolution2. SPOT 6 – 4 bands, 6 m spatial resolution3. Sentinel-2 – 13 bands, 10 m spatial resolution4. EMIT – 285 bands, 60 m spatial resolution Contents Python scripts: Code used for each classification for each imagery . Training data: Field-collected training data points used to train the classifier. Groundtruth data: Field-collected data used to test accuracy and reliability of results.  Classification results: GeoTIFF files showing the classification outputs for each dataset. Metadata document: A detailed explanation of each dataset.   Data accessibily  To ensure replicability, we provide the training datasets and code used in this study. Access to imagery is as follows: Aerial Photographs: Not shared here due to permission restriction. South African users can request data from the Chief Directorate: National Geo-spatial Information (CD: NGI) in Pretoria. SPOT 6: Not shared here due to permission restriction. South African researchers can request data from the South African National Space Agency (SANSA). Sentinel-2: Freely available (for example on Google Earth Engine) EMIT: Can be downloaded from NASA Earthdata: search.earthdata.nasa.gov