Characterization of Volcanic Cloud Components Using Machine Learning Techniques and SEVIRI Infrared Images.

Volcanic explosive eruptions inject several different types of particles and gasses into the atmosphere, giving rise to the formation and propagation of volcanic clouds. These can pose a serious threat to the health of people living near an active volcano and cause damage to air traffic. Many efforts have been devoted to monitor and characterize volcanic clouds. Satellite infrared (IR) sensors have been shown to be well suitable for volcanic cloud monitoring tasks. Here, a machine learning (ML) approach was developed in Google Earth Engine (GEE) to detect a volcanic cloud and to classify its main components using satellite infrared images. We implemented a supervised support vector machine (SVM) algorithm to segment a combination of thermal infrared (TIR) bands acquired by the geostationary MSG-SEVIRI (Meteosat Second Generation-Spinning Enhanced Visible and Infrared Imager). This ML algorithm was applied to some of the paroxysmal explosive events that occurred at Mt. Etna between 2020 and 2022. We found that the ML approach using a combination of TIR bands from the geostationary satellite is very efficient, achieving an accuracy of 0.86, being able to properly detect, track and map automatically volcanic ash clouds in near real-time.

Cell counting and velocity algorithms for hydrodynamic study of unsteady biological flows in micro-channels.

In this paper, the combination of two algorithms, a cell counting algorithm and a velocity algorithm based on a Digital Particle Image Velocimetry (DPIV) method, is presented to study the collective behavior of micro-particles in response to hydrodynamic stimuli. A wide experimental campaign was conducted using micro-particles of different natures and diameters (from 5 to 16 µ m ), such as living cells and silica beads. The biological fluids were injected at the inlet of a micro-channel with an external oscillating flow, and the process was monitored in an investigated area, simultaneously, through a CCD camera and a photo-detector. The proposed data analysis procedure is based on the DPIV-based algorithm to extrapolate the micro-particles velocities and a custom counting algorithm to obtain the instantaneous micro-particles number. The counting algorithm was easily integrated with the DPIV-based algorithm, to automatically run the analysis to different videos and to post-process the results in time and frequency domain. The performed experiments highlight the difference in the micro-particles hydrodynamic responses to external stimuli and the possibility to associate them with the micro-particles physical properties. Furthermore, in order to overcome the hardware and software requirements for the development of a real-time approach, it was also investigated the possibility to detect the flows by photo-detector signals as an alternative to camera acquisition. The photo-detector signals were compared with the velocity trends as a proof of concept for further simplification and speed-up of the data acquisition and analysis. The algorithm flexibility underlines the potential of the proposed methodology to be suitable for real-time detection in embedded systems.