In precision agriculture, pesticides and other inputs shall be used precisely when (and where) they are needed. European Directive 2009/128/EC calls for respecting the principles of integrated pest management (IPM) in the member states. To clarify the question, when, for instance, fungicide use is needed, the well-established economic principle of IPM may be used. This principle says that pests shall be controlled when the costs of control correspond with the damage the pests will cause. Disease levels corresponding with the costs of control are referred to as control thresholds in IPM. Several models have been developed in plant pathology to predict when epidemics will occur, but hardly any of these models predicts a control threshold directly limiting their usefulness for answering the question when pest control is needed according to the principles of IPM. Previously, we quantified the temporal distance between critical rainfall periods and the breaking of the control threshold of Zymoseptoria tritici on winter wheat as being affected by temperature, based on data from 52 field experiments carried out in Luxembourg from 2005 to 2016. This knowledge was used to construct the ShIFT (SeptorIa ForecasT, https://shift.list.lu/) model, which has been validated using external data recorded between 2017 and 2019. Within the efficacy period of a systemic fungicide, the model allowed correct predictions in 84.6% of the cases, while 15.4% of the cases were predicted falsely. The average deviation between the observed and predicted dates of epidemic outbreaks was 0.62 ± 2.4 days with a maximum deviation of 19 days. The observed and predicted dates were closely correlated (r = 0.92, P < 0.0001). Apart from outliers, the forecast model tested here was reliable within the period of efficacy of current commercial fungicides.

Electronically Assisted Astronomy is a fascinating activity requiring suitable conditions and expertise to be fully appreciated. Complex equipment, light pollution around urban areas and lack of contextual information often prevents newcomers from making the most of their observations, restricting the field to a niche expert audience. With recent smart telescopes, amateur and professional astronomers can capture efficiently a large number of images. However, post-hoc verification is still necessary to check whether deep sky objects are visible in the produced images, depending on their magnitude and observation conditions. If this detection can be performed during data acquisition, it would be possible to configure the capture time more precisely. While state-of-the-art works are focused on detection techniques for large surveys produced by professional ground-based observatories, we propose in this paper several Deep Learning approaches to detect celestial targets in images captured with smart telescopes, with a F1-score between 0.4 and 0.62 on test data, and we experimented them during outreach sessions with public in Luxembourg Greater Region.

Modern automated telescopes allow to capture astronomical images in a reproducible way. During the MILAN research project (MachIne Learning for AstroNomy), we have observed deep sky with a Stellina observation station for twelve months from the Luxembourg Greater Region. Thus, we have captured raw images of more than 188 deep sky objects visible from the Northern Hemisphere (galaxies, stars clusters, nebulae, etc.), We have compiled and published this data as the MILAN Sky Survey dataset, allowing interested researchers, industry practitioners and citizens to reuse it.

Solar flares are among the most powerful and dynamic events in the solar system, resulting from the sudden release of magnetic energy stored in the Sun’s atmosphere. These energetic bursts of electromagnetic radiation can release up to ${10}^{32}$ erg of energy, impacting space weather and posing risks to technological infrastructure and therefore require accurate forecasting of solar flare occurrences and intensities. This study evaluates the predictive performance of three machine learning algorithms—Random Forest (RF), k-Nearest Neighbors (kNN), and Extreme Gradient Boosting (XGBoost)—for classifying solar flares into four categories (B, C, M, X). Using 13 parameters of the SHARP dataset, the effectiveness of the models was evaluated in binary and multiclass classification tasks. The analysis utilized 8 principal components (PCs), capturing 95% of data variance, and 100 PCs, capturing 97.5% of variance. Our approach uniquely combines binary and multiclass classification with different levels of dimensionality reduction, an innovative methodology not previously explored in the context of solar flare prediction. Employing a 10-fold stratified cross-validation and grid search for hyperparameter tuning ensured robust model evaluation. Our findings indicate that RF and XGBoost consistently demonstrate strong performance across all metrics, benefiting significantly from increased dimensionality. The insights of this study enhance future research by optimizing dimensionality reduction techniques and informing model selection for astrophysical tasks. By integrating this newly acquired knowledge into future research, more accurate space weather forecasting systems can be developed, along with a deeper understanding of solar physics.