Session CD2 - Ways to improve our space weather forecasting capabilities

Ioannis Daglis, onsite (University of Athens), Stefaan Poedts (KU Leuven), Yuri Shprits (GFZ Potsdam)

In 2019 the European Commission selected three projects for funding under the activity "Horizon-2020 Space: Secure and safe space environment": EUHFORIA 2.0, PAGER and SafeSpace. All three projects focus on ways to improve our space weather forecasting capabilities, covering distinct yet overlapping areas of space weather physics and space weather effects. The session encourages submissions from researchers and advisors working in the framework of these projects, as well as from those working on similar modelling and frameworks.

Poster Viewing
Monday October 24, 09:00 - 14:00, Poster Area

Talks
Tuesday October 25, 13:30 - 14:45, Fire Hall
Tuesday October 25, 15:45 - 16:45, Water Hall
Tuesday October 25, 17:00 - 18:00, Water Hall

Click here to toggle abstract display in the schedule

Talks : Time schedule

Tuesday October 25, 13:30 - 14:45, Fire Hall

13:30	CHRONNOS archive: a comprehensive catalog of solar coronal holes from multi-instrument data	Jarolim, R et al.	Oral
	Robert Jarolim[1], Astrid Veronig[1,2], Stefan Hofmeister[3], Tataiana Podladchikova[4]
	[1]University of Graz; [2]Kanzelhöhe Observatory for Solar and Environmental Research; [3]Leibniz-Institute for Astrophysics Potsdam; [4]Skolkovo Institute of Science and Technology
	Solar coronal holes are a decisive component to study the medium- and long-term evolution of the solar magnetism. As source region of fast solar wind streams, they are critical for the assessment of space weather risks. Space-based observations enable a direct identification of coronal holes from EUV observations. As part of the EU H2020 SOLARNET project we develop AI methods for synoptic full-disk observations of the Sun. In this study, we combine deep learning methods for inter-calibration and automatic coronal hole detection to create a comprehensive and uniform catalog of pixel-wise labeled coronal holes. We use EUV filtergrams from the SDO, STEREO and SOHO mission to automatically identify coronal holes. The catalog features 26 years of full-disk coronal hole maps from 1996 to 2022, includes multiple vantage points, and uses a cadence of up to 6 hours. For the STEREO and SDO observations, we provide a complete summary of the coronal hole magnetic properties. From the multi-viewpoint observations of the STEREO-twin and SDO observations we construct synchronic maps (full-Sun) that allow to study the entire lifetime of coronal holes, without solar rotation induced gaps. We provide an initial statistical evaluation and tools to automatically detect coronal holes for arbitrary times and cadence. With this study we aim to advance the understanding of coronal holes, their evolution, magnetic properties, and relation to the solar cycle, as well as improve space weather forecasting capabilities. In this presentation, we introduce the methodology that enables reliable coronal hole detections, provide an overview of the catalog and present the initial evaluation.
13:45	An inner boundary condition for solar wind models based on coronal density	Bunting, K et al.	Oral
	Kaine Bunting, Huw Morgan
	Aberystwyth University, Aberystwyth University
	Accurate forecasting of the solar wind has grown in importance as society becomes increasingly dependent on technology that is susceptible to space weather events. This work presents a new inner boundary condition based on coronal plasma density gained from tomographic density maps. The density maps provide a direct constraint of the coronal structure at distances where the solar wind is formed and is predominantly radial in flow (4 to 8 solar radii), thus avoiding the complex inner corona which is most difficult to model. An empirical inverse relationship is used to convert densities to solar wind velocities, and this forms the inner boundary condition for the highly-efficient HUXt solar wind model. Comparisons are made between the results gained using the new tomography-based inner boundary condition with results derived using a commonly-used magnetic field extrapolation inner boundary condition. Results show the use of density maps gained from tomography as an inner boundary constraint is a valid alternative to coronal magnetic models and offers a significant advancement in the field given the availability of routine space-based coronagraph observations.
14:00	Modeling of the Earth-directed CME on 2021 October 28	Valentino, A et al.	Oral
	Angelo Valentino[1], Jasmina Magdalenic[1][2]
	[1]Centre for Mathematical Plasma Astrophysics, KU Leuven, Leuven, Belgium , [2]Solar-Terrestrial Center of Excellence-SIDC, Royal Observatory of Belgium, Av. Circulaire 3, B1180 Brussels, Belgium
	We present the study of the solar eruptive event observed on 2021 October 28th. The event originated from the active region NOAA AR 2887, situated in the moment of eruption on the south solar hemisphere (S26 W07). The GOES X1.0 class flare peaked at 15:17 UT, and it was associated with a full halo CME first time observed in the SOHO/LASCO C2 field of view at around 16:00 UT. The CME/flare event was associated with the EIT wave and the coronal dimming. The CME propagated with the 3D speed of approximately 1200 km/s. As the CMEs propagation direction was southward from the Sun-Earth line, only a shock wave was observed by in-situ instruments at 1 AU, in the morning of October 31st. The clearly non radial propagation direction of this CME makes it suitable for our study that focuses on the importance of the propagation direction on the CME arrival time to Earth. We reconstructed the CME using SOHO/LASCO and STEREO/COR observations and modeled it with EUHFORIA and the cone model. The modeled arrival time at Earth was about 10 to 12 hours earlier than the observed one. This early arrival time of the CME was induced by the radial propagation direction of the studied CME. As the cone model assumes purely radial propagation, in order to mimic the non-radial propagation we shifted the latitude of ejection to the south for 15 degrees. All other CME parameters remained the same. The modeling result was significantly improved, with only 11-minutes difference between the observed and the modeled arrival time. These results show the large importance of the direction of propagation when modeling a CME.
14:15	Real-time modelling and forecasting of solar wind disturbances from cradle to Earth	Pinto, R et al.	Oral
	Rui F. Pinto [1,2], R. Kieokaew[1], B. Lavraud[1,3], V. Génot[1], A. Rouillard[1], E. Samara[4], S. Poedts[4], A. Brunet[5], S. Bourdarie[5], Ioannis A. Daglis[6]
	[1] IRAP, Université de Toulouse; UPS-OMP, CNRS; Toulouse, France, [2] 1 LDE3, Université Paris-Saclay, CEA Saclay, France, [3] Laboratoire d’Astrophysique de Bordeaux, Univ. Bordeaux, CNRS, Pessac, France, [4] CmPA, KU Leuven, Belgium & Royal Observatory of Belgium, Brussels, Belgium, [5] ONERA, Toulouse, France, [6] National and Kapodistrian University of Athens, Greece
	We present the solar wind forecast pipeline developed in the scope of the H2020 SafeSpace project, highlighting the part of the pipeline that forecasts the properties of the nascent solar wind (MULTI-VP) and drives models of the solar wind propagation across the heliosphere (HELIO1D and EUHFORIA). The overarching goal of this project is to use several tools in a modular fashion to address the physics of Sun - interplanetary space - Earth's magnetosphere, allowing for comparison to spacecraft measurements and to the formulation of space weather warnings. The solar wind generation model constitutes the first element in chain of models, and is driven by coronal field reconstruction methods using different magnetogram sources (WSO, GONG, ADAPT). Subsequently, the heliospheric propagation models forecast the formation of Corotating Interaction Regions (CIRs) and their high-speed streams (HSS), and will be used to propagate CMEs on the predicted solar wind background. Finally, the outputs of the solar wind and heliosphere models connect to neural network models that predict geomagnetic indices (such as Kp). We take an ensemble forecasting approach to provide optimum forecast up to 2 – 4 days of lead time. Various validation and calibration schemes are introduced at model interfaces in order to select optimal subsets of the ensemble and to correct for model biases, and to potentiate interactions with other space weather actors and services. We will describe the current capabilities of the solar wind forecasting system as well as the future steps in terms of quality control and performance optimisation. This work has received funding from the European Union"s Horizon 2020 research and innovation programme under grant agreement No 870437.
14:30	Visualizing Enlil and EUHFORIA CME Propagation Models using an Accessible Interactive 3-Dimensional Data Visualization Tool	Pankratz, C et al.	Oral
	Christopher Pankratz[1], Greg Lucas[1], Jenny Knuth[1], Thomas E Berger[2]
	[1] Laboratory for Atmospheric and Space Physics (LASP) / Univ. of Colorado, [2] University of Colorado, Space Weather Technology, Research, and Education Center
	Solar wind models such as Enlil and EUHFORIA are critical to informing space weather forecasters of the direction and speed of coronal mass ejections CMEs and informing studies of CME evolution. These models calculate the propagation of the solar wind throughout the 3D heliosphere, but large data volumes and current visualization capabilities have traditionally been restricted to 2D planes intersecting the Earth or other solar system satellites. Here, we present an update on the new open source and publicly accessible visualization capability we have been developing to take advantage of the full 3D data volumes produced by the Enlil model, and more recently the EUHFORIA model. With this tool, a forecaster or other user can interactively visualize the expansion of a CME over time and outside of the plane intersecting the Earth. We have been collaborating closely with researchers and forecasters at the Met Office in the UK and the Space Weather Prediction Center (SWPC) in the USA to develop a tool to enable full view of the heliosphere in a manner that can be tailored to both of these types of users. We have also been collaborating with NASA’s Community Coordinated Modeling Center (CCMC), as well as the developers of both the Enlil and EUHFORIA models in order to connect the visualization tool to the model output from these two models. To enable this capability given the very large output data volumes produced by these models, the visualization interface is accessible via a standard web browser, with both the output data and visualization engine residing together within an Amazon Web Services (AWS) Cloud-based model staging platform computing environment, which permits interactive visualization without the need for a user to transfer large data volumes to their local computer. In this presentation, we will discuss our collaborative engagement with space weather forecasters and researchers worldwide to design this new tailorable interactive 3D visualization tool, and will demonstrate use of the actual visualization tool with different data sets.

Tuesday October 25, 15:45 - 16:45, Water Hall

15:45	Utilizing far-side active regions detected by helioseismology as input to magnetograms for 360° synchronic solar wind forecasting	Heinemann, S et al.	Oral
	Stephan G. Heinemann[1], Dan Yang[1], Jens Pomoell[2]
	[1] Max Planck Institute for Solar System Research, 37077 Göttingen, Germany [2] Department of Physics, University of Helsinki, 00014 Helsinki, Finland
	Synoptic magnetic field data usually serves as the boundary condition for simulations of the global magnetic field; however, it has been shown that these data suffer from an “aging effects” as the longitudinal 360° information can only be obtained over the course of one solar rotation. To avoid this, we use advanced magnetograms produced by feeding near-side HMI/SDO magnetograms and far-side helioseismic active regions into a modified surface flux transport model to improve the modeling of the far-side magnetic configuration. This allows for a more accurate description of the state of the global magnetic field and thus for an improved forecasting of solar wind parameters. We use potential field source surface modeling (PFSS) and the EUropean Heliospheric FORecasting Information Asset (EUHFORIA) to derive the coronal magnetic field configuration and the heliospheric structure as well as discuss the changes caused by the implementation of far-side active regions into magnetic field maps. Modeled solar wind results are found to be in good agreement with far-side in-situ measurements taken by various instruments. We can show the importance of considering not only the solar near-side but also the far-side to accurately model the heliosphere in which solar transients are propagated.
16:00	EUHFORIA simulation using AI generated farside magnetogram	Valliappan, S et al.	Oral
	Senthamizh Pavai Valliappan[1], Jasmina Magdalenic[1,2], Luciano Rodriguez[1]
	[1]Royal Observatory of Belgium, Belgium; [2]Katholieke Universiteit Leuven, Belgium
	EUHFORIA 2.0 (EUropean Heliospheric FORecasting Information Asset, Pomoell & Poedts, 2018) is a space weather prediction tool that simulates the time-evolution of the inner heliospheric plasma environment using a combination of empirical and physics-based modelling approaches. Although a number of studies have been devoted to the simulation of solar wind parameters with EUHFORIA, majority of them aim to distances at about 1 AU, and only few address the close to the Sun distances. With the availability of Parker Solar Probe (PSP), we can test the performance of EUHFORIA at short radial distances. The results of our study of solar wind simulations with EUHFORIA in comparison to the in situ data of PSP, show that using GONG synoptic magnetograms as input to EUHFORIA does not always provide accurate modelling of the solar wind parameters, in particular for the times when the PSP is at the far side of the Sun. These first results were motivation to mitigate the problem by employing more frequently updated magnetograms. Artificial Intelligence (AI)-generated Solar Farside Magnetograms (AISFMs) created using Solar Terrestrial Relations Observatory (STEREO) and Solar Dynamics Observatory (SDO) are now publicly available (Jeong et al., 2022). We employ the AISFMs in EUHFORIA with the aim of achieving better modelling results of solar wind parameters, and at more intervals of PSP positions which helps us in analysing the performance of EUHFORIA at distances closer to the Sun.
16:15	Expected operational solar wind forecast gains from assimilation of in situ L5 observations	Turner, H et al.	Oral
	Harriet Turner[1], Mathew Owens[1], Matthew Owens[1], Siegfried Gonzi[2]
	[1]University of Reading; [2]Met Office
	For accurate space weather forecasting, advanced knowledge of the ambient solar wind is required. Data assimilation (DA) combines model output and observations to form an optimum estimation of reality and is starting to be used in solar wind forecasting. Previous results have shown improvement in solar wind forecasting capabilities when using DA. For DA to be used operationally, it must be able to perform well with real time data. Here, we assimilate distant observations from the Solar Terrestrial Relations Observatory (STEREO) and near-Earth observations from the ACE and DSCOVR spacecraft. We show that forecasts using real time data are not significantly worse than forecasts using cleaned-up, science level data. With a proposed space weather monitor planned for the L5 Lagrange point, we also test the solar wind forecast advantage that observation location can provide.
16:30	Coordinated observations of relativistic and ultra-relativistic electron enhancements following the arrival of consecutive Corotating Interaction Regions	Nasi, A et al.	Oral
	Afroditi Nasi[1], Christos Katsavrias[1], Ioannis A. Daglis[1,2], Ingmar Sandberg[3], Sigiava Aminalragia-Giamini[1,3], Wen Li[4], Yoshizumi Miyoshi[5], Hugh Evans[6], Takefumi Mitani[7], Ayako Matsuoka[8], Iku Shinohara[7], Takeshi Takashima[7], Tomoaki Hori[5], Georgios Balasis[9]
	[1]Department of Physics, National and Kapodistrian University of Athens (NKUA), Athens, Greece; [2]Hellenic Space Center (HSC), Athens, Greece; [3]Space Applications & Research Consultancy (SPARC), Greece; [4]Center for Space Physics, Boston University, Boston, MA, USA; [5]Institute for Space-Earth Environmental Research (ISEE), Nagoya University, Nagoya, Japan; [6]European Space Research and Technology Centre (ESTEC), European Space Agency (ESA), Noordwijk, The Netherlands; [7]Institute of Space and Astronautical Science (ISAS), Japan Aerospace Exploration Agency (JAXA), Japan; [8]Kyoto University, Kyoto, Japan; [9]Institute for Astronomy, Astrophysics, Space Applications and Remote Sensing (IAASARS), National Observatory of Athens, Athens, Greece
	During July to October of 2019, a sequence of Corotating Interaction Regions ($V_{SW}$ ≥ 600 km/s) impacted the magnetosphere, for four consecutive solar rotations, without any interposed Interplanetary Coronal Mass Ejections. Even though the series of CIRs resulted in relatively weak geomagnetic storms ($SYM-H_{min}$ ≈ -60 nT, $Kp_{max}$ ≈ 5), the net effect of the outer radiation belt during each disturbance was different, depending on the electron energy. During the August-September CIR group, significant multi-MeV electron enhancements occurred, up to ultra-relativistic energies of 9.9 MeV in the heart of the outer Van Allen radiation belt. We exploit coordinated data from the Van Allen Probes, THEMIS, Arase and Galileo satellites, to investigate the relative contribution of radial diffusion and gyro-resonant acceleration, investigating the electron fluxes and their Phase Space Density (PSD) profile dependence on the values of the second adiabatic invariant K, ranging from near-equatorial to off equatorial mirroring populations. Additionally, we use chorus wave amplitude (from POES & MetOP satellites) and ULF Pc4-5 power spectral density ($PSD_{B,E}$) estimations from the SafeSpace $D_{LL}$ database (from THEMIS satellites), density measurements from the EMFISIS database (from the Van Allen Probes), as well as solar wind and geomagnetic parameters (from NASA OMNIWeb and SuperMAG databases), for a detailed investigation of these events. Our results indicate that different acceleration mechanisms took place for different electron energies. The PSD profiles were dependent not only on the μ value, but also on the K value, with higher K values corresponding to more pronounced local acceleration by chorus waves. Finally, all ultra-relativistic enhancements took place below geosynchronous orbit, emphasizing the need for more Medium Earth Orbit (MEO) missions. This work has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 870437 for the SafeSpace project.

Tuesday October 25, 17:00 - 18:00, Water Hall

17:00	On the statistics of the radial diffusion coefficients in the outer radiation belt	Katsavrias, C et al.	Oral
	Christos Katsavrias[1], Sigiava Aminalragia-Giamini[1,2], Afroditi Nasi[1], Constantinos Papadimitriou[1,3], Ioannis A. Daglis[1,3], Nourallah Dahmen[4], Antoine Brunet[4], and Sebastien Bourdarie[4]
	[1]Department of Physics, National and Kapodistrian University of Athens, Greece, [2]Space Applications and Research Consultancy (SPARC), Athens, Greece, [3]Hellenic Space Center, Athens, Greece, [4]ONERA, French Aerospace Lab, Toulouse, France
	Radial diffusion has been established as one of the most important mechanisms contributing to the acceleration and loss of relativistic electrons in the outer radiation belt. Therefore, the power spectral density (PSD) of ULF waves, as well as the corresponding radial diffusion coefficients (DLL), are crucial for radiation belt simulations. Over the past few years efforts have been devoted to identify empirical relationships of radial diffusion coefficients (DLL) for radiation belt simulations, yet several studies have suggested that the difference between the various models can be orders of magnitude different at high levels of geomagnetic activity, as the observed DLL have been shown to be highly event-specific. In this work we present a new product of the University of Athens, the ULF PSD and DLL database, which was created in the framework of the EU funded SafeSpace project. We present the statistics on the evolution of DLL during the solar cycle 24 with respect to the various solar wind and geomagnetic parameters, as well as the importance of the use of event-specific DLL through simulations of seed and relativistic electrons with the Salammbo code. This work has received funding from the European Union's Horizon 2020 research and innovation programme “SafeSpace” under grant agreement No 870437.
17:15	Nowcasting radial diffusion coefficients from solar wind: The EMERALD model in the framework of the SafeSpace project	Aminalragia-giamini, S et al.	Oral
	S. Aminalragia-Giamini[1],[2], Christos Katsavrias[1], Constantinos Papadimitriou[1],[2], Ioannis A. Daglis[1],[3], Afroditi Nasi[1], Antoine Brunet[4], Nourallah Dahmen[4], Sebastien Bourdarie[4]
	[1]Department of Physics, National and Kapodistrian University of Athens, Greece, [2]Space Applications and Research Consultancy (SPARC), Athens, Greece, [3]Hellenic Space Center, Athens, Greece, [4]ONERA, French Aerospace Lab, Toulouse, France
	Radial diffusion is one of the dominant physical mechanisms driving acceleration and loss of radiation belt (RB) electrons, thus the estimation of the radial diffusion coefficients (DLL) is of outmost importance for radiation belt models. Over the past years a number of semi-empirical models for have been developed for the estimation of radial diffusion coefficients. However, several studies have suggested that the difference between the outputs of the various models and measurement-derived DLL can be orders of magnitude. To that end, we exploit the DLL database, developed in the framework of the SafeSpace project, to create a DLL nowcasting/forecasting model using Machine Learning methods. We present the model which is used for the derivation of DLL values solely from solar wind parameters and is able to nowcast separately and concurrently the electric and magnetic components (DELL and DBLL respectively) of the radial diffusion coefficient. The model uses probabilistic neural networks which provide expected (mean) estimations along with the respective uncertainties of the estimations. This allows the transition from a deterministic nowcast to a robust probabilistic one which can either provide realistic confidence levels or be integrated in RB simulations. The performance of the model is evaluated and shown along with aspects regarding the successful reproduction of the DLL characteristics. This work has received funding from the European Union's Horizon 2020 research and innovation programme “SafeSpace” under grant agreement No 870437.
17:30	Towards a flexible framework for community-wide forecasting tailored for major space environment impacts	Kuznetsova, M et al.	Oral
	Masha Kuznesova
	NASA Goddard Space Flight Center
	To build continuously improving space weather predictive capabilities based on science and enable assessments and rapid implementations of advances in research into source-to-impact modelling systems we need: to assemble parts of the puzzle by solving problems focused on specific physical domains; to identify essential space environment quantities (ESEQs) passed between domains and linked to impacts; evaluate modeling capabilities for each ESEQ; connect all validated solutions from space weather origins on the sun to impacts on humans and critical infrastructure; to design displays for ensemble predictions tailored for major space weather user groups; to build a collaborative environment for efficient sharing of information and capabilities (models/data/expertise) and collaborative development. The presentation will overview existing community-wide space weather forecasting frameworks and research-to-operations pipelines and discuss opportunities to build a collaborative plug-and-play platform for interconnecting predictive capabilities developed under different space weather programs.
17:45	The ESA Virtual Space Weather Modelling Centre-Part 3	Poedts, S et al.	Oral
	Stefaan Poedts and the VSWMC-P3 team
	KU Leuven
	The goal of the ESA project "Virtual Space Weather Modelling Centre - Part 3" (2019-2021) was to further develop the Virtual Space Weather Modelling Centre, building on the Part 2 prototype system and focusing on the interaction with the ESA SSA SWE system. The objectives and scope of this new project include maintaining the current operational system, the efficient integration of 11 new models and many new model couplings, including daily automated end-to-end (Sun to Earth) simulations, the further development and wider use of the coupling toolkit and front-end GUI, making the operational system more robust and user-friendly. The VSWMC-Part 3 project finished recently. The 11 new models that have been integrated are Wind-Predict (a global coronal model from CEA, France), the Coupled Thermosphere/Ionosphere Plasmasphere (CTIP) model, Multi-VP (another global coronal model form IRAP/CNRS, France), the BIRA Plasma sphere Model of electron density and temperatures inside and outside the plasmasphere coupled with the ionosphere (BPIM, Belgium), the SNRB (also named SNB3GEO) model for electron fluxes at geostationary orbit (covering the GOES 15 energy channels >800keV and >2MeV) and the SNGI geomagnetic indices Kp and Dst models (University of Sheffield, UK), the SPARX Solar Energetic Particles transport model (University of Central Lancashire, UK), Spenvis DICTAT tool for s/c internal charging analysis (BISA, Belgium), the Gorgon magnetosphere model (ICL, UK), and the Drag Temperature Model (DTM) and operations-focused whole atmosphere model MCM being developed in the H2020 project SWAMI. Many new couplings have also been implemented and a dynamic coupling facility has been installed. Moreover, Daily runs are implemented of two model chains covering the whole Sun-to-Earth domain. The results of these daily runs are made available to all VSWMC users. We will provide an overview of the state-of-the-art, including the new available model couplings and daily model chain runs, and demonstrate the system. VSWMC-P3 team: Alexis Rouillard, Andrea Lani, Andrey Kochanov, Antoine Strugarek, Barbara Perri, Daniel Heynderickx, David Berghmans, Jonathan Eastwood, Edith Botek, Erwin Dedonder, Fabian Diet , Francois Boschet, Jan Depauw, Jan Ooghe, Jesse Andries, Johan De keyzer, Luciano Rodriguez, Mag Selwa, Michail A. Balikhin, Nicolae Mihalache, Norma Crosby, Paul Borgermans, Petra Vanlommel, Richard Boynton, Robbe Vansintjan, Rui Pinto, Sacha Brun, Silvia Dalla, Simo

Posters

1	Radial diffusion coefficients dependence on ICME and SIR driven disturbances	Thanasoula, K et al.	Poster
	Konstantina Thanasoula[1], Christos Katsavrias[1], Afroditi Nasi[1], Ioannis A. Daglis[1],[2], Georgios Balasis[3], and Theodore Sarris[4]
	[1]Department of Physics, National and Kapodistrian University of Athens, Greece, [2]Institute for Accelerating Systems and Applications (IASA), Athens, Greece, [3]IAASARS, National Observatory of Athens, Greece, [4]Department of Electrical Engineering, Democritus University of Thrace, Greece
	Radial diffusion driven by Ultra Low Frequency (ULF) waves is very important for magnetospheric dynamics, because it contributes to relativistic electron enhancements and losses in the outer Van Allen radiation belt. The dependence of ULF wave power spectral density and radial diffusion coefficients (DLL) on solar wind parameters has already been investigated. In this study, we use the "SafeSpace" database (https://synergasia.uoa.gr/modules/document/?course=PHYS120), which includes radial diffusion coefficients DLL and ULF wave power spectral density. This database was created using magnetic and electric field measurements by the THEMIS satellites for a 9-year period (2011- 2019). We conduct a statistical analysis of radial diffusion coefficients DLL, which contributes to relativistic electron radial diffusion quantification, to find out their dependence on interplanetary drivers (25 events of Interplanetary Coronal Mass Ejections (ICMEs) and 46 events of Stream Interaction Regions (SIRs) of which the 16 events are SIRs with shock and the 30 events are SIRs without shock). We study how the parameters of these solar wind drivers influence the behavior of radial diffusion coefficients. The results indicate significant differences between ICME and SIR driven disturbances at the ratio of DLLE to DLLB, especially when the solar wind dynamic pressure is maximized. This feature introduces a significant energy dependence to the radial diffusion coefficients, which is further depending on the radial distance and the different mu values.
2	An assessment of the performance of the EUHFORIA2.0 chain of models, from Sun to Earth, in predicting GIC in power grids across Europe	Clarke, E et al.	Poster
	Ellen Clarke[1], Ewelina Florczak[1], Guanren Wang[1], Ciarán Beggan[1], Gemma Richardson[1], Alan Thomson[1], Aurélie Marchaudon[2], Pierre-Louis Blelly[2], Julian Eisenbeis[2], Simon Thomas[2], Jimmy Raeder[3], Banafsheh Ferdousi[3], Anwesha Maharana[4] and Stefaan Poedts[4]
	[1]British Geological Survey; [2]Institut de Recherche en Astrophysique et Planétologie; [3]University of New Hampshire; [4]KU Leuven
	One aim of the unique chain of upgraded EUHFORIA2.0 coupled models is the prediction of geomagnetically induced current (GIC) across Europe and to provide a means to analyse space weather events on power grids. The work reported here specifically examines the performance of the section of the coupled chain from L1 to GIC. Magnetospheric and ionospheric models predicting ground level variations in the magnetic field are combined with ground conductivity models to compute geo-electric fields across Europe. These, in turn, drive the power grid models that estimate GIC at any given node on the modelled grid. The focus is not on the individual models, the details of which are presented elsewhere at this meeting, but on the analysis of the capability to predict GIC in power grids with a greater lead time than has been attempted previously. The full coupled chain has been used for selected historical geomagnetic storms and forecast accuracy has been examined by comparing model results with measured magnetic field variations at observatories and networks across Europe. GIC predictions have also been compared to any available measured GIC. Quantitative, summary statistics and metrics are provided and a qualitative overview of the results presented. This work is funded by the European Commission Horizon 2020 (H2020) Grant Agreement No. 870405.
3	Prediction of Adverse effects of Geomagnetic storms and Energetic Radiation (PAGER)	Shprits, Y et al.	Poster
	Yuri Shprits[1], Hayley Allison[1], Dedong Wang[1], Michael Wutzig[1], Stefano Bianco[1], Ruggero Vasile[1], Bernhard Haas[1], Tony Arber[2], Keith Bennett[2], Mike Liemohn[3], Bart van der Holst[3], Ondrej Santolik[4], Ivana Kolmasova[4], Ulrich Taubenschuss[4], Julien Forest[5], Arnaud Trouche[5], Benoît Tezenas du Montcel[5].
	[1]GFZ German Research Centre for Geosciences, Germany [2]Department of Physics, University of Warwick, UK, [3]Climate and Space Sciences and Engineering, University of Michigan, USA, [4]Institute of Atmospheric Physics, Czech Academy of Sciences, CZ, [5]Artenum SARL, FR
	The PAGER project aims to provide space weather predictions that will be initiated from observations on the Sun and to predict radiation in space and its effects on satellite infrastructure. Real-time predictions and a historical record of the dynamics of the cold plasma density and ring current allow for evaluation of surface charging, and predictions of the relativistic electron fluxes will allow for the evaluation of deep dielectric charging. The project aims to provide a 1-2 day probabilistic forecast of ring current and radiation belt environments, which will allow satellite operators to respond to predictions that present a significant threat. As a backbone of the project, we use the most advanced codes that currently exist and adapt existing codes to perform ensemble simulations and uncertainty quantifications. This project includes a number of innovative tools including data assimilation and uncertainty quantification, new models of near-Earth electromagnetic wave environment, ensemble predictions of solar wind parameters at L 1, and data-driven forecast of the geomangetic Kp index and plasma density. The developed codes may be used in the future for realistic modelling of extreme space weather events. The PAGER consortium is made up of leading academic and industry experts in space weather research, space physics, empirical data modelling, and space environment effects on spacecraft from Europe and the US.
4	Impact of high speed solar wind streams on the dynamic variations of the electron population in the outer Van Allen belt.	Triantopoulou, A et al.	Poster
	Alexandra Triantopoulou[1], Afroditi Nasi[1], Christos Katsavrias[1], Ingmar Sandberg [2] and Ioannis A. Daglis[1],[3]
	[1] Department of Physics, National and Kapodistrian University of Athens, Athens, Greece, [2] Space Applications & Research Consultancy (SPARC), Athens, Greece, [3] Institute for Accelerating Systems and Applications (IASA), Athens, Greece
	More than half a century after the discovery of Van Allen radiation belts, their dynamics are still not fully understood. The variability of the outer radiation belt electrons is related to geospace disturbances due to the interactions of solar wind with the terrestrial magnetosphere. During such interactions the relativistic and ultra-relativistic electron population can be enhanced, depleted or have no significant change depending on the relative contribution of the various acceleration and loss mechanisms. This variability depends significantly on the solar wind driver type. One of the most common solar wind drivers that affect, amongst others, the electrons of the outer radiation belt is the High Speed Streams (HSSs) following Stream Interaction Regions (SIRs). Their effect on the various energetic electron populations is a major theme of space research. In this work, we examine the electron phase space density (PSD) during 45 SIR-driven events that occurred during the RBSP era (2012-2018), in terms of superposed epoch analysis (SEA), for seed, relativistic and ultra-relativistic electrons (μ=100, 1000, 5000 MeV/G). Furthermore, we examine three distinct values of the second adiabatic invariant (K) in order to compare the differences between local acceleration and radial diffusion. In addition, we selected the time of magnetopause maximum compression (Lmpmin) as the zero epoch time for all events, with the aim of extracting a statistical study of the response of the outer belt electron population on this solar wind driver.
5	Status and future plans of the UK SWIMMR SPF programme	Burnett, C et al.	Poster
	Dr Ian McCrea, Catherine Burnett
	RAL Space, Science & Technology Facilities Council (STFC)
	SWIMMR (Space Weather Innovation, Measurement, Modelling and Risk) is a £20M, five-year programme of space weather capacity building for the UK, funded by the Strategic Priorities Fund of UK Research and Innovation. It involves eleven different projects supported by two UK research councils (STFC and NERC) and supported by three government ministries (BEIS, Ministry of Defence and the Department for Transport) with the Met Office Space Weather Operations Centre as the major beneficiary. The programme is now in its fourth year and the first models are being delivered to the UK Met Office, with a number of new instruments either under construction or already deployed. This presentation will give an update on the current status of the SWIMMR activities set out the plans for the remaining period of the programme and summarise the most recent plans for activities beyond the current programme.
6	A Synthetic Test of Ensemble Forecasting of the Energetic Particle Fluxes at Geosynchronous Orbit	Liu, H et al.	Poster
	Ruotong Liu, Jian Yang
	Southern Univ. of Science and Technology
	It has long been recognized that one critical component of space weather forecasting is the prediction of energetic particle fluxes along geosynchronous orbit. However, the particle fluxes may vary by orders of magnitudes over time and space. The dynamic mechanisms of the variations are found to be quite complex. Especially, the tens of KeV particle fluxes at nightside sector are related to substorm injections. Therefore, the prediction of fluxes presents scientists with a very difficult challenge. Ensemble forecasting has been used in numerical weather prediction for many years as a way to combine different predictions given the specific weather conditions. In this paper, an ensemble forecasting method is exploited to deduce a linear combination model that can be used to predict the high energetic particle fluxes at geosynchronous orbit solely based on a set of simulation results from the Rice Convection Model. This study illustrates the capability of the ensemble technique to improve particle fluxes forecasts in the synthetic test.
7	Impact of Interplanetary Coronal Mass Ejections on the dynamic variations of the electron population in the outer Van Allen belt	Dimitrakoula, A et al.	Poster
	Adamantia Dimitrakoula[1], Afroditi Nasi[1], Christos Katsavrias[1], Ingmar Sandberg[2] and Ioannis A. Daglis[1][3]
	[1]Department of Physics, National and Kapodistrian University of Athens, Athens, Greece, [2]Space Applications & Research Consultancy (SPARC), Athens, Greece, [3]Institute for Accelerating Systems and Applications (IASA), Athens, Greece
	The outer Van Allen radiation belt is an environment with intense variability due to complex mechanisms that are part of the solar–terrestrial coupling. A fundamentally important effect is the acceleration of seed electrons to relativistic and ultra-relativistic energies. In our work, we examine 19 events from the Van Allen Probes era (2012 – 2018), which are chosen according to the interplanetary driver of the geomagnetic disturbance. The events are caused by Interplanetary Coronal Mass Ejections (ICMEs), for which we calculate the electron Phase Space Density (PSD) for distinct values of the first adiabatic invariant (μ=100,1000,5000 MeV/G) corresponding to the different energies of the electrons in the outer radiation belt. Furthermore, we study the response of the electron population at different K values (second adiabatic invariant), in order to compare the differences between the two main acceleration mechanisms and their dependence on the electron pitch angle. This is achieved by performing a Superposed Epoch Analysis (SEA) of the geomagnetic disturbance events, taking into consideration the parameters of solar wind and the state of the magnetosphere. We compare and discuss the variability of the electron PSD for different values of the adiabatic invariants, the effects of the ICMEs on the outer radiation belt and how the different electron populations are affected.
8	Physics-based machine learning for CMEs forecasting	Candiani, V et al.	Poster
	Valentina Candiani [1], Sabrina Guastavino [1], Francesco Marchetti [2], Alessandro Bemporad [3], Roberto Susino [3], Daniele Telloni [3], Anna Maria Massone [1], Michele Piana [1]
	[1] University of Genova, [2] University of Padova, [3] Istituto Nazionale di AstroFisica - Osservatorio Astrofisico di Torino
	Coronal mass ejections (CMEs) are among the largest eruptions in the solar system, and the main cause for intense and energetic geomagnetic storms. Fast and accurate prediction of CME arrival time is thus essential to minimize the disruption that CMEs may cause when reaching the Earth. In this work, we propose a new approach for CME arrival time prediction using physics-based machine learning techniques, while exploiting the well-established drag-based model. To this end, the CME parameters used in our analysis are the CME mass, initial speed, cross-section area, and the solar wind speed and density. These properties are inferred from orbital satellites’ coronagraphs such as LASCO (Large Angle and Spectrometric Coronagraph). We addressed the prediction problem with a two-step method. First we estimate the drag coefficient and then we train a neural network by optimizing a loss function composed of two different parts: a classical data-driven mismatch term and a novel physics-based loss term incorporating the drag-based model forward solution. Preliminary results show that such technique may be a valuable tool for the CME arrival time prediction. Further validation of the presented two-step model with larger sets of data is still underway.
9	An open platform for validating solar wind model solutions at Earth	Reiss, M et al.	Poster
	Martin A. Reiss[1,2], Karin Muglach[3], Richard Mullinix[1], Maria M. Kuznetsova[1], Chiu Wiegand[1], and the Ambient Solar Wind Validation Team (H1-01)
	[1]Community Coordinated Modeling Center, NASA GSFC, Greenbelt, MD, USA; [2]Austrian Space Weather Office, Graz, Austria; [3]NASA Goddard Space Flight Center, Greenbelt, MD, USA,
	Progress in modeling the large-scale corona and inner heliosphere needs community-wide strategies and procedures to evaluate the abilities of our modeling assets. We present the progress of the Ambient Solar Wind Validation Team embedded in the COSPAR ISWAT initiative. Our team's mission is to provide the science community with an assessment of the state-of-the-art in solar wind forecasting at Earth and other planetary environments. To this end, we are developing an open online platform hosted at NASA's CCMC for validating solar wind models by comparing their solutions with in situ spacecraft measurements. The online platform will allow the space weather community to test the quality of state-of-the-art solar wind models with unified metrics providing an unbiased assessment of uncertainties in model solutions. In this contribution, we will give a status update on our team effort, showcase a first version of the platform, and outline future perspectives.
10	Space Weather Landscape in Slovakia	Mackovjak, S et al.	Poster
	Simon Mackovjak [1]
	[1] Institute of Experimental Physics, Slovak Academy of Sciences
	Slovak National Space Safety Program (SK-S2P) study is a current activity supported by European Space Agency (ESA). Its goal is to identify Slovak capabilities for S2P topics, map these capabilities to the needs of research organizations and the space industry in ESA member states, and prepare a roadmap for future growth of S2P expertise in Slovakia. During the contribution, the key points, main capabilities, and prepared roadmap for future development in the Space Weather domain in Slovakia will be presented.
11	An Improved Lifetime Model for the High Energy Electrons in the Near-Earth Space Due to Their Interactions With Chorus Waves	Wang, D et al.	Poster
	Dedong Wang[1], Yuri Shprits [1,2,3], Bernhard Haas[1,2]
	[1]GFZ German Research Centre for Geosciences, Potsdam, Germany; [2]University of Potsdam, Potsdam, Germany; [3]University of California, Los Angeles, California, USA
	Improving forecasting capabilities of the high-energy electrons in near-Earth space is an important task in our space weather community. Wave-particle interactions play a critical role in the dynamics of these electrons. Whistler mode chorus waves can cause both acceleration and loss of these electrons. To quantify the dynamics of these energetic electrons caused by chorus waves, we developed a chorus wave model by using several years of NASA’s Van Allen Probe data and extended the model to higher latitudes by referring to the measurements from ESA’s Cluster mission. Using this chorus wave model, we calculated diffusion coefficients to quantify the interactions between chorus waves and energetic electrons. Then we performed long-term simulations with diffusion in three dimensions in the phase space for the whole Van Allen Probe era. We carefully validated our simulation results against the electron observations and found they agree well. However, for some codes that include many physical processes, it is very expensive to include 3D diffusions to account for wave-particle interactions in their simulations. Thus, based on the diffusion coefficients we used in our simulations, we provide our community with an improved lifetime model to account for the loss caused by chorus waves. In this study, we parameterize the lifetime of the electrons with an energy range from 1 keV to 2 MeV. In each magnetic local time (MLT), we calculated the lifetime for each energy and L-shell using two different methods. By applying polynomial fits, we parameterized the electron lifetime as a function of L-shell and electron kinetic energy (Ek) in each MLT and geomagnetic activity (Kp). During storm time, the lifetimes for higher energy (> keV) electrons range from hours to days in the heart of the radiation belts. In contrast, the lifetimes for electrons with lower energy (< 100 keV) range from minutes to hours. The current model has much better MLT coverage compared with previous lifetime models. It is currently being used in two different magnetosphere codes and the results show better agreement with satellite measurements. We share our improved electron lifetime model with our community and we are happy to collaborate with people who are interested in it. This work has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No. 870452 for the PAGER project.
12	SafeSpace: a radiation belt forecasting project for the safety of space assets	Daglis, I et al.	Poster
	Ioannis A. Daglis [1], Sebastien Bourdarie [2], Stefaan Poedts [3], Ondrej Santolik [4], Fabien Darrouzet [5], Juan Cueto Rodriguez [6], Benoit Lavraud [7], Ingmar Sandberg [8], Christos Katsavrias [1], Afroditi Nasi [1], George Balasis [1], Omiros Giannakis [1], Konstantina Moutsouroufi [1], Stefanos Doulfis [1], Marina Georgiou [1], Fiori-Anastasia Metallinou [1], Antoine Brunet [2], Nourallah Dahmen [2], Vincent Maget [2], Evangelia Samara [3], Benjamin Grison [4], Ivana Kolmasova [4], David Pisa [4], Jan Soucek [4], Viviane Pierrard [5], Edith Botek 5], Ion Bueno Ulacia [6], Jose Manuel Jimenez Cerezo [6], Gaizka Eiguren Arza [6], Jesus Angel Oliveros Fernandez [6], Luis de Pablo [6], Rui Pinto [7], Rungployphan (Om) Kieokaew [7], Vincent Genot [7], Constantinos Papadimitriou [8], Sigiava Aminalragia-Giamini [8], and Zafar Iqbal [8]
	[1] Department of Physics, National and Kapodistrian University of Athens, Athens, Greece, [2] ONERA/DESP, Toulouse, France, [3] Centre for Plasma Astrophysics - KU Leuven, Leuven, Belgium, [4] Institute of Atmospheric Physics, Prague, Czechia, [5] Belgian Institute for Space Aeronomy, Brussels, Belgium, [6] Thales Alenia Space España, Madrid, Spain, [7] IRAP/University of Toulouse-France/CNRS, Toulouse, France, [8] SPARC - Space Applications & Research Consultancy, Athens, Greece
	The European SafeSpace project has been implementing a synergistical approach to improve space weather forecasting capabilities from the current lead times of a few hours to 2-4 days. We have combined the solar wind acceleration model MULTI-VP with the heliospheric propagation models Helio1D and EUHFORIA to compute the evolution of the solar wind from the surface of the Sun to the Earth orbit. The forecasted solar wind conditions are then fed into the ONERA Geoffectiveness Neural Network, to forecast the level of geomagnetic activity with the Kp index as the chosen proxy. The Kp index is used as the input parameter for the IASB plasmasphere model, which is used to estimate VLF wave amplitude and then VLF diffusion coefficients, while the predicted solar wind parameters are used to estimate the ULF diffusion coefficients – both critical components of the integrated model, as wave-particle interactions are a key factor for the improved radiation belt predictions. Plasmaspheric density and VLF/ULF diffusion coefficients are used by the Salammbô radiation belts code to deliver a detailed flux map of energetic electrons. Another key characteristic of the SafeSpace integrated model is the improved estimation and propagation of uncertainties, which is particularly useful for the Salammbô code, and allows for more precise assimilation of in-situ measurements. Ultimately, electron radiation indicators are also provided as a prototype space weather service, accessible at http://www.safespace-service.eu. The performance of the prototype service has been evaluated in collaboration with space industry stakeholders. The work leading to this paper has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 870437 for the SafeSpace (Radiation Belt Environmental Indicators for the Safety of Space Assets) project.