Ocean surface monitoring, emphasizing oil slick detection, has become essential due to its importance for oil exploration and ecosystem risk prevention. Automation is now mandatory since the manual annotation process of oil by photo-interpreters is time-consuming and cannot process the data collected continuously by the available spaceborne sensors. Studies on automatic detection methods mainly focus on Synthetic Aperture Radar (SAR) data exclusively to detect anthropogenic (spills) or natural (seeps) oil slicks, all using limited datasets. The main goal is to maximize the detection of oil slicks of both natures while being robust to other phenomena that generate false alarms, called “lookalikes”. To this end, this paper presents the automation of offshore oil slick detection on an extensive database of real and recent oil slick monitoring scenarios, including both types of slicks. It relies on slick annotations performed by expert photo-interpreters on Sentinel-1 SAR data over four years and three areas worldwide. In addition, contextual data such as wind estimates and infrastructure positions are included in the database as they are relevant data for oil detection. The contributions of this paper are: (i) A comparative study of deep learning approaches using SAR data. A semantic and instance segmentation analysis via FC-DenseNet and Mask R-CNN, respectively. (ii) A proposal for Fuse-FC-DenseNet, an extension of FC-DenseNet that fuses heterogeneous SAR and wind speed data for enhanced oil slick segmentation. (iii) An improved set of evaluation metrics dedicated to the task that considers contextual information. (iv) A visual explanation of deep learning predictions based on the SHapley Additive exPlanation (SHAP) method adapted to semantic segmentation. The proposed approach yields a detection performance of up to $94\%$ of good detection with a false alarm reduction ranging from $14\%$ to $34\%$ compared to mono-modal models. These results provide new solutions to improve the detection of natural and anthropogenic oil slicks by providing tools that allow photo-interpreters to work more efficiently on a wide range of marine surfaces to be monitored worldwide. Such a tool will accelerate the oil slick detection task to keep up with the continuous sensor acquisition. This upstream work will allow us to study its possible integration into an industrial production pipeline. In addition, a prediction explanation is proposed, which can be integrated as a step to identify the appropriate methodology for presenting the predictions to the experts and understanding the obtained predictions and their sensitivity to contextual information. Thus it helps them to optimize their way of working.

When rockfalls hit and damage linear stakes such as roads or railways, the access to critical infrastructures (hospitals, schools, factories …) might be disturbed or stopped. Rockfall risk management often involves building protective structures that are traditionally based on the intensive use of resources such as steel or concrete. However, these solutions are expensive, considering their construction and maintenance, and it is very difficult to protect long linear stakes. A more sustainable and effective risk management strategy could be to account for changes on rockfall activity related to weather conditions. By integrating sustainability principles, we can implement mitigation measures that are less resource-intensive and more adaptable to environmental changes. For instance, instead of solely relying on physical barriers, solutions could include measures such as restriction of access, monitoring and mobilization of emergency kits containing eco-friendly materials. A critical step in developing such a strategy is accurately predicting periods of increased rockfall activity according to meteorological triggers. In this paper, we test four machine learning models to predict rockfalls on the National Road 1 at La Réunion, a key road for the socio-economic life of the island. Rainfall and rockfall data are used as inputs of the predictive models. We show that a set of features derived from the rainfall and rockfall data can predict rockfall with performances very close and almost slightly better than the standard expert model used for operational management. Metrics describing the performance of these models are translated in operational terms, such as road safety or the duration of road closings and openings, providing actionable insights for sustainable risk management practices.