Search

Artificial intelligence (AI) methods have emerged as a powerful tool to study and in some cases forecast natural disasters [1,2]. Recent works have successfully combined deep learning modeling with scientific knowledge stemming from the SAR Interferometry domain propelling research on tasks like volcanic activity monitoring [3], associated with ground deformation. A milestone in this interdisciplinary field has been the release of the Hephaestus [4] InSAR dataset, facilitating automatic InSAR interpretation, volcanic activity localization as well as the detection and categorization of atmospheric contributions in wrapped interferograms. Hephaestus contains annotations for approximately 20,000 InSAR frames, covering the 44 most active volcanoes in the world. The annotation was performed by a team of InSAR experts that manually examined each InSAR frame individually. However, even with such a large dataset, class imbalance remains a challenge, i.e. the InSAR samples containing volcano deformation fringes are orders of magnitude less than those that do not. This is anticipated since natural hazards are in principle rare in nature. To counter that, the authors of Hephaestus provide more than 100,000 unlabeled InSAR frames to be used for global large-scale self-supervised learning, which is more robust to class imbalance when compared to supervised learning [5]. Motivated by the Hephaestus dataset and the insights provided by [2], we train global, task-agnostic models in a self-supervised learning fashion that can handle distribution shifts caused by spatio-temporal variability as well as major class imbalances. By finetuning such a model to the labeled part of Hephaestus we obtain the backbone for a global volcanic activity alerting system, namely Pluto. Pluto is a novel end-to-end AI based system that provides early warnings of volcanic unrest on a global scale.Pluto automatically synchronizes its database with the Comet-LiCS [6] portal to receive newly generated Sentinel-1 InSAR data acquired over volcanic areas. The new samples are fed to our volcanic activity detection model. If volcanic activity is detected, an automatic email is sent to the service users, which contains information about the intensity, the exact location and the type (Mogi, Sill, Dyk) of the event. To ensure a robust and ever-improving service we augment Pluto with an iterative pipeline that collects samples that were misclassified in production, and uses them to further improve the existing model. [1] Kondylatos et al. "Wildfire danger prediction and understanding with Deep Learning." Geophysical Research Letters 49.17 (2022): e2022GL099368.[2] Bountos et al. "Self-supervised contrastive learning for volcanic unrest detection." IEEE Geoscience and Remote Sensing Letters 19 (2021): 1-5.[3] Bountos et al. "Learning from Synthetic InSAR with Vision Transformers: The case of volcanic unrest detection." IEEE Transactions on Geoscience and Remote Sensing (2022).[4] Bountos et al. "Hephaestus: A large scale multitask dataset towards InSAR understanding." Proceedings of the IEEE/CVF CVPR. 2022.[5] Liu et al. "Self-supervised learning is more robust to dataset imbalance." arXiv preprint arXiv:2110.05025 (2021).[6] Lazecký et al. "LiCSAR: An automatic InSAR tool for measuring and monitoring tectonic and volcanic activity." Remote Sensing 12.15 (2020): 2430.

The 15 January 2022 eruption of the Hunga volcano provides a unique opportunity to study the reaction of the ionosphere to large explosive events. In particular, this event allows us to study the global propagation of travelling ionospheric disturbances (TIDs) using various instruments. We focus on detecting the ionospheric disturbances caused by this eruption over Europe, where dense networks of both ionosondes and GNSS receivers are available. This event took place on the day of a geomagnetic storm. We show how data from different instruments and observatories can be combined to distinguish the TIDs produced by the eruption from those caused by concurrent geomagnetic activity. The Lamb wavefront was detected as the strongest disturbance in the ionosphere, travelling between 300 and 340 m/s, consistent with the disturbances in the lower atmosphere. By comparing observations obtained from multiple types of instruments, we also show that TIDs produced by various mechanisms are present simultaneously, with different types of waves affecting different physical quantities. This illustrates the importance of analysing data from multiple independent instruments in order to obtain a full picture of an event like this one, as relying on only a single data source might result in some effects going unobserved.

Flood has become the most frequent and deadliest type of disaster by far, responsible for the 43.5% of deaths in 2019. What is more, the number of flood events has extremely increased during the last decade (2000-2019), compared to the previous one (1980-1999) (CRED 2020). Therefore, policy and decision makers, more than ever, need efficient flood monitoring tools in order to facilitate their work towards increasing disaster resilience, especially in the urban and peri-urban areas, where most of the population and critical infrastructure are located. For this purpose, the FloodHub system has been developed by the Center of Earth Observation and Satellite Remote Sensing BEYOND, at the National Observatory of Athens, in the framework of the EuroGEO Disaster Resilience Action Group, supported by on-going actions (SMURBS / ERA-PLANET and Excelsior H2020 projects and the sponsor Hellenic Petroleum S.A.). The innovation of the system lies in the integration of different data sources, so as to deliver a reliable flood early warning system, and an operational awareness picture of the crisis every 5’ to the relevant authorities, namely on three levels: municipality, region, and national civil protection. FloodHub allows the near-real-time ingestion and assimilation of hydrometeorological measurements from in-situ telemetric stations, Sentinels data, and crowdsourced data, in a multi-source data fusion concept, using sophisticated hydrologic and hydraulic modelling and statistical regression techniques. It offers increased reliability through a continuous validation and optimization of results, automation in assimilating flood modeling in real time, computational efficiency, openness, flexibility, scalability, transferability, and the speed to meet rapid awareness during the crisis. Therefore, FloodHub is a useful tool in the hands of the relevant authorities and key stakeholders, contributing to an effective flood risk and crisis management. This is in line with the requirements for the implementation of the EU Floods Directive 2007/60/EC, the Sendai Framework for Disaster Risk Reduction, the UN SDGs, as well as the GEO’s Societal Benefit Areas.

Big, open Earth Observation data that are currently made available by space agencies are not following the linked data paradigm. Therefore, from the perspective of a user, the Earth Observation data and other kinds of geospatial data necessary to satisfy his or her information need can only be found in different data silos, where each silo may contain only part of the needed data. Opening up these silos by publishing their contents as RDF and interlinking them with semantic connections allows the development of data analytics applications with great environmental and financial value. In this paper, we present the life-cycle of big, linked and open Earth Observation data and we show how to support its various stages using the software stack developed by the projects EU research projects TELEIOS and LEO. We also show this stack of tools can be used to implement a wild fire monitoring service.

The purpose of this paper is to evaluate the performance of the TSMP-assisted Digisonde (TaD) topside profiling technique. We present systematic comparisons between electron density profiles and TEC parameters extracted from TaD model with (a) CHAMP-derived TEC parameters, (b) CHAMP reconstructed profiles, (c) ground based GPSderived TEC parameters, and (d) profiles reconstructed from RPI/IMAGE plasmagrams. In all cases, TaD follows the general trend of plasmaspheric observations derived from the above datasets. Especially during storm cases, TaD shows remarkable agreement with the variations of the ground based GPS-derived TEC parameters. Overall, the comparison results shows that TaD method can be adopted by EURIPOS to provide the electron density distribution up to plasmaspheric heights in real-time.

[1] Ionospheric drift measurements are of significant importance in the monitoring of the upper atmosphere dynamics and an important contributor toward the deeper understanding of the ionospheric electrodynamics. An interesting perspective for operational ionospheric drift measurements comes from the HF radars, such as the Digisondes. Nevertheless, when interpreting velocity measurements obtained from the Digisonde drift analysis method, care must be taken in relating the motion to the effects of photoionization, recombination, gravity waves, and plasma motion. In this paper we investigate the effect of all different mechanisms contributing to the apparent velocity measured by the Athens Digisonde. As a first step, the daily drift pattern over Athens under quiet magnetospheric conditions is preliminarily determined for the equinox months. Moreover, the substorm effect on the extracted pattern is also investigated. Our results provide some evidence for the direct effect of traveling atmospheric disturbances in the vertical component, while in the horizontal plane, the disturbance could be attributed to other substorm-related mechanisms.

This study aims at providing experimental evidence, to support the hypothesis according to which the movement of the moon's shadow sweeping the ozone layer at supersonic speed, during a solar eclipse, creates gravity waves in the atmosphere. An experiment was conducted to study eclipse induced thermal fluctuations in the ozone layer (via measurements of total ozone column, ozone photolysis rates and UV irradiance), the ionosphere (Ionosonde Total Electron Content – ITEC, peak electron density height – hmF2), and the troposphere (temperature, relative humidity), before, during and after the total solar eclipse of 29 March 2006. We found the existence of eclipse induced dominant oscillations in the parameters related to the ozone layer and the ionosphere, with periods ranging between 30–40 min. Cross-spectrum analyses resulted to statistically significant square coherences between the observed oscillations, strengthening thermal stratospheric ozone forcing as the main mechanism for GWs. Additional support for a source below the ionosphere was provided by the amplitude of the oscillations in the ionospheric electron density, which increased upwards from 160 to 220 km height. Even though similar oscillations were shown in surface temperature and relative humidity data, no clear evidence for tropospheric influence could be derived from this study, due to the modest amplitude of these waves and the manifold rationale inside the boundary layer.