Background Ailanthus altissima is one of the worst invasive plants in Europe. It reproduces both by seeds and asexually through root sprouting. The winged seeds can be dispersed by wind, water and machinery, while its robust root system can generate numerous suckers and cloned plants. In this way, Ailanthus altissima typically occurs in very dense clumps, but can also occasionally grow as widely spaced or single stems. This highly invasive plant can colonise a wide range of anthropogenic and natural sites, from stony and sterile soils to rich alluvial bottoms. Due to its vigour, rapid growth, tolerance, adaptability and lack of natural enemies, it spreads spontaneously, out-competing other plants and inhibiting their growth Introduction Over the last few decades, Ailanthus altissima has quickly spread in the Alta Murgia National Park (Southern Italy) which is mostly characterized by dry grassland and pseudo-steppe, wide-open spaces with low vegetation, which are very vulnerable to invasion. Ailanthus altissima causes serious direct and indirect damages to ecosystems, replacing and altering communities that have great conservation value, producing severe ecological, environmental and economic effects, and causing natural habitat loss and degradation. The spread of Ailanthus altissima is likely to increase in the future, unless robust action is taken at all levels to control its expansion. In a recent working document of the European Commission, it was found that the cost of controlling and eliminating invasive species in Europe amounts to €12 billion per year. Two relevant questions then arise: i) whether it is possible or not to fully eradicate or, at least, to reduce the impact of an invasive species and ii) how to achieve this at a minimum cost, in terms of both environmental damage and economic resources. A Life Program funded the Life Alta Murgia project (LIFE12BIO/IT/000213) had, as its main objective, the eradication of this invasive exotic tree species from the Alta Murgia National Park. That project provided both the expert knowledge and valuable in-field data for the Ailanthus validation case study, which was conceived and developed within the Internal Joint Initiative of LifeWatch ERIC. Aims At the start of the on-going eradication program a single map of A. altissima was available, dating back to 2012. Due to the lack of data, predicting the extent of invasion and its impacts was extremely difficult, making it impossible to assess the efficacy of control measures. Static models based on statistics cannot predict spatial–temporal dynamics (e.g. where and when A. altissima may repopulate an area), whereas mechanistic models incorporating the growth and spread of a plant would require precise parametrisation, which was extremely difficult with the scarce information available. To overcome these limitations, a relatively simple mechanistic model has been developed, a diffusion model, which is validated against the current spatial distribution of the plant estimated by satellite images. This model accounts for the effect of eradication programs by using a reaction term to estimate the uncertainty of the prediction. This model provides an automatic tool to estimate a-priori the effectiveness of a planned control action under temporal and budget constraints. This robust tool can be easily applied to other geographical areas and, potentially, to different species.

Output file from the "1st stage Data-driven Classifier" service (step 4) of the Ailanthus Workflow within the Internal Joint Initiative: confusion matrix related to the validation dataset. File format: text file.

Output file from the "2nd stage Data-driven Classifier" service (step 9) of the Ailanthus Workflow within the Internal Joint Initiative: confusion matrix related to the validation dataset. File format: text file.

Input file to the "1st stage Data-driven Classifier" service (step 4) of the Ailanthus Workflow within the Internal Joint Initiative. Parameters to be setted for the training of the SVM classifier: Kernel function; Regularization parameter; penalty parameter. File format: text file.

Input file to the "2nd stage Data-driven Classifier" service (step 9) for the 2-classes classification of the Ailanthus Workflow within the Internal Joint Initiative. Parameters to be setted for the training of the SVM classifier: Kernel function; Regularization parameter; penalty parameter. File format: text file.

Output file from the "1st stage Data-driven Classifier" service (step 4) and input file to the "Extractor Resampler and Masking" service (step 8) of the Ailanthus Workflow within the Internal Joint Initiative: look-up table to identify the correspondence between numeric and FAO-LCCS codes. File format: text file.

Output file from the "2nd stage Image Stacking" service (step 5) and the input file to the "Extractor Resampler and Masking" service (step 8) of the Ailanthus Workflow within the Internal Joint Initiative. It is a multi-layer file containing the two multi-season images each one with its 8 bands. File format: raster geotiff.

Output file from the "1st stage Splitter" service (step 2) and input file to the "1st stage Data-driven Classifier" service (step 4) of the Ailanthus Workflow within the Internal Joint Initiative. It represents the reference data for the validation of the output map obtained from the supervised multi-class data-driven classifier. There is a file for each class in the scene labelled according to FAO-LCCS taxonomy. A numeric code is associated to each FAO_LCCS label. File format: shapefile.

Input file to the "1st stage Splitter" service of the Ailanthus Workflow within the Internal Joint Initiative. It contains the reference data for all the classes in the scene, as collected in-field, labelled according to FAO-LCCS taxonomy. These data will be splitted into multiple files (each one for each land cover class) to be used for the training of the supervised data-driven classifier. File format: shapefile.

Input file to the "1st stage Splitter" service (step 2) of the Ailanthus Workflow within the Internal Joint Initiative. It contains the reference data for all the classes in the scene, as collected in-field, labelled according to FAO-LCCS taxonomy. These data will be splitted into multiple files (each one for each land cover class) to be used for the validation of the output map obtained from the supervised data-driven classifier. File format: shapefile.