TY - CONF AU - Lemperger, István AU - Erdős, Géza AU - Márkusné Bebesi, Zsófia AU - Kovács, Péter Gábor AU - Németh, Zoltán AU - Molnár, Csaba AU - Kuslits, Lukács AU - Kis, Árpád AU - Dwivedi, Navin Kumar AU - Szalai, Sándor AU - Wesztergom, Viktor TI - Tuning of low magnetic field in the focal volume of the ZBL T2 - IAGA / IASPEI Joint Scientific Meeting 2025 PY - 2025 PG - 1 UR - https://m2.mtmt.hu/api/publication/36859096 ID - 36859096 N1 - Poster LA - English DB - MTMT ER - TY - CONF AU - Lemperger, István AU - Erdős, Géza AU - Márkusné Bebesi, Zsófia AU - Kovács, Péter Gábor AU - Németh, Zoltán AU - Molnár, Csaba AU - Kis, Árpád AU - Dwivedi, Navin Kumar AU - Szalai, Sándor AU - Kuslits, Lukács AU - Wesztergom, Viktor TI - Establishment of the Zero Magnetic Field Laboratory (ZBL) at the Széchenyi István Geophysical Observatory (SzIGO) T2 - IAGA / IASPEI Joint Scientific Meeting 2025 PY - 2025 PG - 1 UR - https://m2.mtmt.hu/api/publication/36859150 ID - 36859150 N1 - Presentation LA - English DB - MTMT ER - TY - CONF AU - Czirok, Lili AU - Kuslits, Lukács AU - Laksona, FX Anjar Tri AU - Bozsó, István AU - Gribovszki, Katalin Eszter TI - Demonstration of clustering-based stress field analyses: insights from Vrancea zone and the 2023 Turkey earthquake sequence T2 - IAGA / IASPEI Joint Scientific Meeting 2025, 31 August - 5 September 2025, Lisbon, Portugal: Book of Abstracts PY - 2025 UR - https://m2.mtmt.hu/api/publication/36921511 ID - 36921511 LA - English DB - MTMT ER - TY - GEN AU - Herein, Mátyás AU - Kuslits, Lukács AU - Jánosi, Dániel TI - Reversal Statistics in the Non-Autonomous Rikitake Dynamo Model PY - 2025 PG - 10 UR - https://m2.mtmt.hu/api/publication/37068417 ID - 37068417 LA - English DB - MTMT ER - TY - CONF AU - Kuslits, Lukács AU - András, Horváth AU - Ciaran, Beggan AU - Rubóczki, Tibor AU - Prácser, Ernő AU - Czirok, Lili AU - Bozsó, István AU - Lemperger, István TI - Physics informed neural networks at the service of geodynamo research T2 - IAGA / IASPEI Joint Scientific Meeting 2025, 31 August - 5 September 2025, Lisbon, Portugal: Book of Abstracts PY - 2025 SP - 98 EP - 98 PG - 1 UR - https://m2.mtmt.hu/api/publication/37068994 ID - 37068994 LA - English DB - MTMT ER - TY - CONF AU - Lemperger, István AU - Géza, Erdős AU - Zsófia, Bebesi AU - Péter, Kovács AU - Németh, Zoltán AU - Molnár, Csaba AU - Kuslits, Lukács AU - Kis, Árpád AU - Dwivedi, Navin Kumar AU - Szalai, Sándor AU - Wesztergom, Viktor TI - Automatic fine tuning of the low magnetic field in the focal space of the ZBL T2 - 46-th COSPAR Scientific Assewmbly Abstract PY - 2025 SP - 1 UR - https://m2.mtmt.hu/api/publication/37070514 ID - 37070514 LA - English DB - MTMT ER - TY - JOUR AU - Kuslits, Lukács AU - Horváth, András AU - Wesztergom, Viktor AU - Beggan, Ciaran AU - Rubóczki, Tibor AU - Prácser, Ernő AU - Czirok, Lili AU - Bozsó, István AU - Lemperger, István TI - An alternate representation of the geomagnetic core field obtained using machine learning JF - EARTH PLANETS AND SPACE J2 - EARTH PLANETS SPACE VL - 76 PY - 2024 IS - 1 PG - 41 SN - 1343-8832 DO - 10.1186/s40623-024-02024-5 UR - https://m2.mtmt.hu/api/publication/34938287 ID - 34938287 AB - Machine learning (ML) as a tool is rapidly emerging in various branches of contemporary geophysical research. To date, however, rarely has it been applied specifically for the study of Earth’s internal magnetic field and the geodynamo. Prevailing methods currently used in inferring the characteristic properties and the probable time evolution of the geodynamo are mostly based on reduced representations of magnetohydrodynamics (MHD). This study introduces a new inference method, referred to as Current Loop-based UNet Model Segmentation Inference (CLUMSI). Its long-term goal focuses on uncovering concentrations of electric current densities inside the core as the direct sources of the magnetic field itself, rather than computing the fluid motion using MHD. CLUMSI relies on simplified models in which equivalent current loops represent electric current systems emerging in turbulent geodynamo simulations. Various configurations of such loop models are utilized to produce synthetic magnetic field and secular variation (SV) maps computed at the core–mantle boundary (CMB). The resulting maps are then presented as training samples to an image-processing neural network designed specifically for solving image segmentation problems. This network essentially learns to infer the parameters and configuration of the loops in each model based on the corresponding CMB maps. In addition, with the help of the Domain Adversarial Training of Neural Networks (DANN) method during training, historical geomagnetic field data could also be considered alongside the synthetic samples. This implementation can increase the likelihood that a network trained primarily on synthetic data will appropriately handle real inputs. Our results focus mainly on the method's feasibility when applied to synthetic data and the quality of these inferences. A single evaluation of the trained network can recover the overall distribution of loop parameters with reasonable accuracy. To better represent conditions in the outer core, the study also proposes a computationally feasible process to account for magnetic diffusion and the corresponding induced currents in the loop models. However, the quality of the reconstruction of magnetic field properties is compromised by occasional poor inferences, and an inability to recover realistic SV. LA - English DB - MTMT ER - TY - JOUR AU - Patkó, Levente AU - Kuslits, Lukács AU - Czirok, Lili AU - Liptai, Nóra TI - Cluster analysis on a suite of upper mantle xenoliths from the Nógrád–Gömör Volcanic Field (Northern Pannonian Basin) JF - FÖLDTANI KÖZLÖNY J2 - FÖLDTANI KÖZLÖNY VL - 154 PY - 2024 IS - 3 SP - 329 EP - 346 PG - 18 SN - 0015-542X DO - 10.23928/foldt.kozl.2024.154.3.329 UR - https://m2.mtmt.hu/api/publication/35467820 ID - 35467820 N1 - HUN-REN Institute of Earth Physics and Space Science, Sopron, Hungary Oscope Research Group, HUN-REN Institute of Earth Physics and Space Science, Sopron, Hungary Quantectum Earthquake Forecasting Center, Ljubljana, Slovenia Export Date: 17 April 2025; Cited By: 0; Correspondence Address: L. Patkó; HUN-REN Institute of Earth Physics and Space Science, Sopron, Hungary; email: patko.levente@epss.hun-ren.hu AB - The Nógrád–Gömör Volcanic Field is one of the most thoroughly studied upper mantle xenolith-bearing localities in the Carpathian–Pannonian region. Dozens of upper mantle xenoliths collected there have been investigated from various aspects in the past decades. Major element compositions of upper mantle xenoliths available in scientific papers on the Nógrád–Gömör xenoliths were extracted to build a database. This database includes the geochemical composition of the rock-forming minerals (olivine, pyroxenes, spinel) of 112 upper mantle xenoliths. Using the data compiled in this database, we applied the CluStress algorithm to perform cluster analysis. The results led to the division of 2 supergroups (I and II) and 17 groups within them. Supergroup I was split into two subgroups (Supergroup I/a and I/b). The xenoliths of Supergroup I/a and I/b are distinguished by their orthopyroxene/clinopyroxene modal composition ratios (mainly >1.5 and <1, respectively) as well as their olivine Mg-numbers (dominantly >0.89 and <0.89, respectively). Most of the xenoliths (~50% of all classified xenoliths) were sorted into Group 1 within Supergroup I/a, and they appear at almost all xenolith-bearing localities in the Nógrád–Gömör. The Group 1 xenoliths exhibit narrow compositional ranges regarding their major element relationships. Due to their common occurence and spatial distribution, these rocks were considered to represent an ambient lithospheric mantle under­neath the study area. The geochemical characteristics of the other groups allowed their linking to different events described in former scientific works. At least 11 out of the 17 distinguished groups (i.e. ~35% of the classified xenoliths) are strongly related to the Neogene alkali basaltic volcanism. This suggests that the pre-Neogene rootzones of volcanic fields in the Carpathian–Pannonian region were significantly modified by intensive melt-rock reactions between upward migrating basaltic melts and wall rocks, which took place in the last ~10 million years. LA - English DB - MTMT ER - TY - GEN AU - Kuslits, Lukács AU - Horváth, András AU - Wesztergom, Viktor AU - Ciaran, Beggan AU - Rubóczki, Tibor AU - Prácser, Ernő AU - Czirok, Lili AU - Bozsó, István AU - Lemperger, István TI - Szintetikus adatok nagy sebességű generálása és képfeldolgozó neurális hálók a földmágnesség kutatásának szolgálatában PY - 2024 UR - https://m2.mtmt.hu/api/publication/35748211 ID - 35748211 AB - A földi mágneses mező alakulásának megértésében évtizedek óta egyre jelentősebb szerep jut a nagy teljesítményű számítási infrastruktúrán végzett numerikus szimulációknak. Ezt részben a mezőt bolygónkban fenntartó folyamat összetettsége, részben annak léptékbeni sajátosságai indokolják. Egy, a folyamat belső viszonyait rekonstruálni képes gépi tanuló algoritmus megfelelő működéséhez szükséges, nagy számosságú és jellegében is kellő változatosságot eredményező számítássorozat generálása azonban épp ezen okoknál fogva jelent komoly kihívást. Az itt bemutatott módszertani fejlesztés a direkt numerikus szimulációk megkerülésével, egy idealizált analitikus modell és egy szisztematikus modellszámítás eredményeinek közelítése révén kísérli meg a fenti problémát orvosolni. A fejlesztés tágabb relevanciáját adhatják például az olyan műszaki vagy orvosdiagnosztikai problémák, melyek esetében nem lehetséges vagy nem kívánatos a vizsgált fizikai folyamathoz közvetlenül hozzáférni, annak komplex modellezése pedig jelentős számításigényű. Az eljárás során alkalmazott tanítási módszer képessé teheti a gépi tanuló (mélytanuló) algoritmusokat arra is, hogy kezeljék a szimulációk és a valós folyamat között feszülő ellentmondásokat. LA - Hungarian DB - MTMT ER - TY - CONF AU - Bozóki, Tamás AU - Młynarczyk, Janusz AU - Bór, József AU - Kubisz, Jerzy AU - Bozsó, István AU - Horváth, András AU - Kuslits, Lukács AU - Timkó, Máté TI - ELF-transients detected in the broadband recordings at the Hylaty station in Poland T2 - EGU General Assembly 2023: Abstracts PB - European Geosciences Union (EGU) C1 - München PY - 2023 DO - 10.5194/egusphere-egu23-17102 UR - https://m2.mtmt.hu/api/publication/33797498 ID - 33797498 AB - Lightning acts as a natural antenna radiating electromagnetic (EM) waves in a wide frequency range. In the extremely low frequency (ELF) band (3 Hz - 3 kHz), lightning-induced EM waves suffer very weak attenuation while they propagate in the waveguide formed by the Earth’s surface and the lowest part of the ionosphere. These EM waves can travel around the Earth several times before losing most of their energy. This allows ELF-transients generated by powerful lightning discharges from around the globe to be detected at any observation site. We developed an algorithm that identifies ELF-transients in the broadband recordings at Hylaty, Poland (sampling frequency: 3004.81 Hz, antenna bandwidth: 0.02 Hz to 1.1 kHz) and finds their most probable source lightning discharge in the lightning database of the Word Wide Lightning Location Network (WWLLN) based on the technique described by Bór et al. (2022).Between July 2020 and April 2021 about 270,000 ELF-transients were found in the records from Hylaty. The most probable source of 160,000 transients was identified in the WWLLN database. Using this data set, we show that the propagation speed of broadband ELF-transients differ significantly when the propagation path is on the dayside or on the nightside of the Earth. It is also demonstrated that for lightning discharges close to Hylaty (d<2Mm), the timing and location accuracy of WWLLN has a large impact on the identification of the lightning source and on the inferred propagation speed. A convolutional neural network, trained with ELF-transients of known source location, was used to determine the distance to the lightning source in cases where the source lightning discharge could not be found in the WWLLN database. The average accuracy of the distance provided by the neural network is 700 km. No significant difference can be seen between the distribution of distances obtained by matching the source lightning stroke in the WWLLN database and that obtained using the neural network-based approach. Reference:Bór, J., Szabóné André, K., Bozóki, T., Mlynarczyk, J., Steinbach, P., Novák, A., and Lemperger, I. (2022): Estimating the Attenuation of ELF-Band Radio Waves in the Earth’s Crust by Q-Bursts. IEEE Transactions on Antennas and Propagation, 70, 8. https://doi.org/10.1109/TAP.2022.3161504 LA - English DB - MTMT ER -