Alvar Aalto 1.0 β! i.e. AiLanthus Visual Analysis and Recognition based on Ailanthus Altissima Labeled Training Objects! Summary Alvar Aalto is a cloud-based application for the reconnaissance and mapping of Ailanthus altissima. The system combines a GIS-prepared training and control dataset (Aalto) with the Google Earth Engine processing script (Alvar), which retrieves multispectral and thermal satellite imagery, extracts key spectral and topographic features, and performs large-scale classification over an area of interest using established machine-learning methods. The output consists of a classified raster map and vectorized polygons of potential A. altissima presence, supporting the assessment of tree invasion patterns.! The Alvar algorithm is based on the assumption that Sentinel-2 imagery, although only barely sufficient in resolution, can still be exploited to detect the presence of such trees. Alvar Aalto is the result of the shared work of the authors, who are each involved in studies related to A. altissima invasiveness. ! Authors Marco Ciolfi 1* ORCiD 0000-0003-4831-8053! Emilio Badalamenti 2 3 ORCiD 0000-0002-8178-354X! Francesca Chiocchini 1 ORCiD 0000-0002-5122-8756! Marco Lauteri 1 ORCiD 0000-0003-1071-7999! Paola Pollegioni 1 3 ORCiD 0000-0001-6388-1931! 1 Institute of Research on Terrestrial Ecosystems of the National Research Council, Porano, Italy! 2 Department of Agricultural, Food and Forest Sciences, University of Palermo, Italy! 3 National Biodiversity Future Center, Palermo, Italy ! * Contact Marco Ciolfi -
[email protected]! Distribution The Aalto training set refers to Pantelleria Island, in the southern Mediterranean, Italy. The Ailanthus altissima presence and absence points were collected through local field surveys. The dataset is released under the Creative Commons Attribution International license CC BY 4.0. ! The Alvar 1.0 β script is distributed under the GNU GPL 3.0 license. It is provided without any warranty of accuracy or performance, and no specific results are claimed or guaranteed.! The distribution package is composed of the following four files:! •_README.pdf – this document! •man.pdf – the manual, covering input preparation and script execution! •alvar.txt – the Google Earth Engine JavaScript code! •aalto.gpkg – a GeoPackage containing an example input dataset based on a field survey on Pantelleria Island, southern Mediterranean, Italy. ! Alvar Aalto: Any reference to the outstanding Finnish architect is purely coincidental; the acronym emerged naturally from its technical components—though the name was admittedly too elegant to resist. doi:10.5281/zenodo.17672767! November 2025
