Foivos Karakostas
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Foivos Karakostas has a total of 234 co-authors. Here are the top co-authors by number of shared publications:
- Philippe Lognonné
Université Paris Cité Institut de Physique du Globe de Paris CNRS Paris France
15 shared publications - Nicholas Schmerr
Department of Geology University of Maryland College Park MD USA
12 shared publications - William Bruce Banerdt
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, USA
10 shared publications - Domenico Giardini
ETH Zurich
8 shared publications - Simon Christian Stähler
Institute of Geophysics ETH Zürich Zürich Switzerland
7 shared publications - Vedran Lekic
Department of Geology, University of Maryland, College Park, MD, USA
6 shared publications
Latest Publications
- 2 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Analysis and modeling of meteor impact and airburst generated seismic waves on terrestrial planets with atmosphere
Analyse et modélisation des ondes sismiques générées par des impacts et des explosions atmosphériques des météores aux planètes telluriques avec une atmosphère: Les évènements météoriques constituent une source d’importance fondamentale pour la sismologie planétaire, étant donné que leur localisation, et dans certains cas, leur temps d’origine peuvent être déterminés précisément par des orbiteurs. Cette importance augmente encore dans le cas d’une expérimentation à 1 seul sismomètre, comme dans le cas de SEIS (Seismic Experimentof Interior Structure) de la mission actuelle InSight (Interior Exploration using Seismic Investigations, Geodesy and Heat Transport). En effet, la localisation précise permet de réaliser une inversion directe des temps de propagation différentiels et des formes d’ondes pour la détermination de la structure interne. Les impacts de météorites génèrent des ondes de volume et de surface lors de leur arrivée à la surface d’une planète. Quand ils explosent dans l’atmosphère, ils produisent des ondes de chocs qui sont converties en ondes linéaires, sismiques pour la partie solide, et acoustiques pour la partie atmosphérique. Ce phénomène peut être modélisé par l’amplitude de l’excitation de modes sphéroïdaux, dû aux effets de couplage entre l’atmosphère et le sol. Ce manuscrit de thèse est consacré à l’investigation et la modélisation des ondes de Rayleigh générées par des météores. Un rappel général des avancées en sciences planétaires est d’abord fourni, avec un focus sur la sismologie planétaire et les études des sources sismiques atmosphériques. Ensuite, la théorie concernant les ondes de choc dans l’atmosphère et au sol est présentée plus en détails. Dans le cas de la formation d’une onde de choc dans l’atmosphère, l’effet de transition d’un régime de propagation non linéaire vers un régime linéaire est documenté pour le superbolide de Chelyabinsk. Pour la génération d’ondes dans la subsurface, un impact de météorite sur la lune est passé en revue, en utilisant des codes hydrodynamiques. Une analyse comparée de ces deux cas est réalisée de façon à présenter les processus de transition du régime de propagation. Une inversion de la source sismique du superbolide de Chelyabinsk est effectuée, de manière à examiner les propriétés de la source associée dans l’atmosphère terrestre. Nous avons développé une source multiple, composée d’une série de points source consécutifs, basé sur une trajectoire fournie. Les calculs des sismogrammes synthétiques des ondes de Rayleigh associées à l’événement sont réalisés par la sommation des modes propres du modèle de la partie solide et de la partie atmosphérique de la planète. A travers une technique d’inversion basée sur la décomposition des valeurs singulières et la méthode du moindre carré, nous fournissons des solutions pour la magnitude du moment. De plus, nous avons trouvé dans les données sismiques un effet Doppler, associée à la directivité de la source. En plus, nous avons réalisé des modélisations 3-D basées sur la méthode des éléments spectraux dans le cas d’un modèle solide uniquement, de façon à comprendre les effets des caractéristiques 3-D crustales, et surligner les différences avec une source inversée dans le sol par rapport à une source correctement positionnée dans l’atmosphère. Dans le cas de Mars, la sommation des modes propres est utilisée pour fournir les formes d’ondes associées aux impacts à la surface de la planète ou à basse altitude dans l’atmosphère martienne. Il est montré que la contribution du mode solide sphéroïdal fondamental domine les formes d’onde, par rapport aux deux premières harmoniques. La comparaison entre les amplitudes de sismogramme synthétiques de tailles différentes, montre que les petits impacteurs (diamètre de 0,5 mètre à 2 mètres) peuvent être détectés par les capteurs VBB de SEIS, seulement pour les hautes fréquences des ondes de Rayleigh, même pour des distances épicentrales très faibles.
- 2 April 2026
- Published by Agence Bibliographique de l'Enseignement Supérieur
Analysis and modeling of meteor impact and airburst generated seismic waves on terrestrial planets with atmosphere
Analyse et modélisation des ondes sismiques générées par des impacts et des explosions atmosphériques des météores aux planètes telluriques avec une atmosphère: Les évènements météoriques constituent une source d’importance fondamentale pour la sismologie planétaire, étant donné que leur localisation, et dans certains cas, leur temps d’origine peuvent être déterminés précisément par des orbiteurs. Cette importance augmente encore dans le cas d’une expérimentation à 1 seul sismomètre, comme dans le cas de SEIS (Seismic Experimentof Interior Structure) de la mission actuelle InSight (Interior Exploration using Seismic Investigations, Geodesy and Heat Transport). En effet, la localisation précise permet de réaliser une inversion directe des temps de propagation différentiels et des formes d’ondes pour la détermination de la structure interne. Les impacts de météorites génèrent des ondes de volume et de surface lors de leur arrivée à la surface d’une planète. Quand ils explosent dans l’atmosphère, ils produisent des ondes de chocs qui sont converties en ondes linéaires, sismiques pour la partie solide, et acoustiques pour la partie atmosphérique. Ce phénomène peut être modélisé par l’amplitude de l’excitation de modes sphéroïdaux, dû aux effets de couplage entre l’atmosphère et le sol. Ce manuscrit de thèse est consacré à l’investigation et la modélisation des ondes de Rayleigh générées par des météores. Un rappel général des avancées en sciences planétaires est d’abord fourni, avec un focus sur la sismologie planétaire et les études des sources sismiques atmosphériques. Ensuite, la théorie concernant les ondes de choc dans l’atmosphère et au sol est présentée plus en détails. Dans le cas de la formation d’une onde de choc dans l’atmosphère, l’effet de transition d’un régime de propagation non linéaire vers un régime linéaire est documenté pour le superbolide de Chelyabinsk. Pour la génération d’ondes dans la subsurface, un impact de météorite sur la lune est passé en revue, en utilisant des codes hydrodynamiques. Une analyse comparée de ces deux cas est réalisée de façon à présenter les processus de transition du régime de propagation. Une inversion de la source sismique du superbolide de Chelyabinsk est effectuée, de manière à examiner les propriétés de la source associée dans l’atmosphère terrestre. Nous avons développé une source multiple, composée d’une série de points source consécutifs, basé sur une trajectoire fournie. Les calculs des sismogrammes synthétiques des ondes de Rayleigh associées à l’événement sont réalisés par la sommation des modes propres du modèle de la partie solide et de la partie atmosphérique de la planète. A travers une technique d’inversion basée sur la décomposition des valeurs singulières et la méthode du moindre carré, nous fournissons des solutions pour la magnitude du moment. De plus, nous avons trouvé dans les données sismiques un effet Doppler, associée à la directivité de la source. En plus, nous avons réalisé des modélisations 3-D basées sur la méthode des éléments spectraux dans le cas d’un modèle solide uniquement, de façon à comprendre les effets des caractéristiques 3-D crustales, et surligner les différences avec une source inversée dans le sol par rapport à une source correctement positionnée dans l’atmosphère. Dans le cas de Mars, la sommation des modes propres est utilisée pour fournir les formes d’ondes associées aux impacts à la surface de la planète ou à basse altitude dans l’atmosphère martienne. Il est montré que la contribution du mode solide sphéroïdal fondamental domine les formes d’onde, par rapport aux deux premières harmoniques. La comparaison entre les amplitudes de sismogramme synthétiques de tailles différentes, montre que les petits impacteurs (diamètre de 0,5 mètre à 2 mètres) peuvent être détectés par les capteurs VBB de SEIS, seulement pour les hautes fréquences des ondes de Rayleigh, même pour des distances épicentrales très faibles.
- Open Access
- 2 May 2024
- Published by Copernicus GmbH
An analysis of the seismic scattering on Mars, using the InSight seismic data
The scattering of seismic waves is the signature of random heterogeneities, present in the lithospheric structure of a terrestrial planet. It is the result of refraction and reflection of the seismic waves generated by a quake, when they cross materials with different shear rigidity, bulk modulus, and density and therefore different seismic wave velocities, compared to the ambient space. On Earth, the seismic waves show relatively weak scattering, identified in later arriving coda waves that follow the main arrivals of body waves and decay with time. In contrast, seismic wave scattering is much more significant on the Moon, where the high heterogeneous structure of the lunar megaregolith, produced through millions of years of impact bombardment, is a structure that creates an extreme scattering environment.The landing of the NASA InSight mission on Mars in 2018, which carried and deployed a seismometer for the first time on the Martian ground, offered a pristine dataset for the investigation and analysis of the characteristics of the scattering attenuation of the Martian crust and uppermost mantle which is important for understanding the structure of the Martian interior. Lognonné et al. (2020) used a methodology based on the radiative transfer model (Margerin et al., 1998) to offer the first constraints for the scattering and attenuation in the Martian crust. In this study, we performed a further examination based on more and newer events of the Martian Seismic Catalog (InSight Marsquake Service, 2021).The Marsquake Catalog contains events that are categorized according to the frequency content of the seismic signal (Clinton et al., 2021). In this study, we used 19 events of 4 different families, namely the Low Frequency, Broadband, High Frequency, and Very High Frequency events, for our investigation. We focused our investigation on the characteristics of the S-coda waveforms and for this reason, we worked on the respective energy envelopes. We manually picked the envelopes, defining the time window of the S-coda waves, as well the frequency range for each event, directly from the spectrograms of the events' signals, using an appropriately developed visual tool.We used a modeling approach (Dainty et al., 1974) that was developed for the computation of the energy envelopes of shallow events (Lunar impacts) and a diffusive, highly scattering layer, sitting over an elastic half-space. The energy envelope depends on the thickness of the diffusive layer, the range of the seismic ray, the diffusivity and the attenuation in the top layer, and the seismic wave velocity in the underneath elastic half-space. We analyzed all the tradeoffs between the terms of the modeling equation, namely the geometrical relationship of the velocity contrast between the diffusive layer and the elastic half-space with the seismic ray range and the diffusive layer thickness, the diffusivity with the diffusive layer thickness, and between the diffusivity and the velocity contrast of the two examined layers.The presence of the aforementioned tradeoffs made the definition of a unique model a very hard task, as the information for the azimuthal characteristics of the signal is not available for the examined events. This is a limitation that exists in seismology only while working with one station, with the InSight seismometer being the only station on a planet, and the amplitude of the seismic signal is not big enough to perform a specific polarization analysis and derive the azimuthal origin of the recorded signal. For this reason, we reviewed the fit between the modeling and the data, depending on the frequency content of the events.The Low Frequency and the Broadband events, which have a frequency content mainly below the tick noise detected at 1 Hz, could not satisfy the modeling approach of a simple diffusive layer. The spectral envelopes of the S-coda waves of these events are decaying very rapidly, which suggests an origin in a more elastic environment. This is in agreement with previous studies (Giardini et al., 2020) that suggest that these events are generated deeper in the Martian mantle. For this reason, we applied another approach to these signals, with an energy envelope equation designed for deep moonquakes (Dainty et al.,1974), but it was not either capable to fit the examined data envelopes, suggesting the absence of a very thick megaregolith structure on Mars.Based on the results of the High Frequency (HF) and Very High Frequency (VF) events we observed a range of possible paths and diffusivities that can satisfy the data and we investigated the tradeoffs between the parameters of a modeling equation that control the shape of the energy envelope for the events. The analysis of these tradeoffs does not permit us to make any assumptions about the depth of the diffusive region in the Martian crust and the upper mantle as their azimuthal characteristics are unknown and therefore it is not feasible to tell if the difference in the result analysis reflects vertical or lateral variations of the uppermost diffusive layer in the Martian lithosphere.The results of this study illustrate one of the challenges in working with single-station seismic data where event location information, including distance, azimuth, and depth are crucial for understanding the lateral variation in seismic properties of a planet. The existence of a seismic network on the planetary scale will improve the ability of phase peaking and location identification of the events and therefore it will give additional constraints for a similar analysis.ReferencesClinton, J. F. et al. (2021). The Marsquake catalogue from InSight, sols 0–478.Phys Earth Planet In, 310:106595.Dainty, A. M. et al. (1974). Seismic scattering and shallow structure of the Moon in Oceanus Procellarum.The Moon,9(1-2):11–29.Giardini, D. et al. (2020). The seismicity of Mars.Nat Geosci, 13(3):205–212.InSight Marsquake Service (2021). Mars Seismic...
- 2 May 2024
- Published by Copernicus GmbH
Open AccessAn analysis of the seismic scattering on Mars, using the InSight seismic data
The scattering of seismic waves is the signature of random heterogeneities, present in the lithospheric structure of a terrestrial planet. It is the result of refraction and reflection of the seismic waves generated by a quake, when they cross materials with different shear rigidity, bulk modulus, and density and therefore different seismic wave velocities, compared to the ambient space. On Earth, the seismic waves show relatively weak scattering, identified in later arriving coda waves that follow the main arrivals of body waves and decay with time. In contrast, seismic wave scattering is much more significant on the Moon, where the high heterogeneous structure of the lunar megaregolith, produced through millions of years of impact bombardment, is a structure that creates an extreme scattering environment.The landing of the NASA InSight mission on Mars in 2018, which carried and deployed a seismometer for the first time on the Martian ground, offered a pristine dataset for the investigation and analysis of the characteristics of the scattering attenuation of the Martian crust and uppermost mantle which is important for understanding the structure of the Martian interior. Lognonné et al. (2020) used a methodology based on the radiative transfer model (Margerin et al., 1998) to offer the first constraints for the scattering and attenuation in the Martian crust. In this study, we performed a further examination based on more and newer events of the Martian Seismic Catalog (InSight Marsquake Service, 2021).The Marsquake Catalog contains events that are categorized according to the frequency content of the seismic signal (Clinton et al., 2021). In this study, we used 19 events of 4 different families, namely the Low Frequency, Broadband, High Frequency, and Very High Frequency events, for our investigation. We focused our investigation on the characteristics of the S-coda waveforms and for this reason, we worked on the respective energy envelopes. We manually picked the envelopes, defining the time window of the S-coda waves, as well the frequency range for each event, directly from the spectrograms of the events' signals, using an appropriately developed visual tool.We used a modeling approach (Dainty et al., 1974) that was developed for the computation of the energy envelopes of shallow events (Lunar impacts) and a diffusive, highly scattering layer, sitting over an elastic half-space. The energy envelope depends on the thickness of the diffusive layer, the range of the seismic ray, the diffusivity and the attenuation in the top layer, and the seismic wave velocity in the underneath elastic half-space. We analyzed all the tradeoffs between the terms of the modeling equation, namely the geometrical relationship of the velocity contrast between the diffusive layer and the elastic half-space with the seismic ray range and the diffusive layer thickness, the diffusivity with the diffusive layer thickness, and between the diffusivity and the velocity contrast of the two examined layers.The presence of the aforementioned tradeoffs made the definition of a unique model a very hard task, as the information for the azimuthal characteristics of the signal is not available for the examined events. This is a limitation that exists in seismology only while working with one station, with the InSight seismometer being the only station on a planet, and the amplitude of the seismic signal is not big enough to perform a specific polarization analysis and derive the azimuthal origin of the recorded signal. For this reason, we reviewed the fit between the modeling and the data, depending on the frequency content of the events.The Low Frequency and the Broadband events, which have a frequency content mainly below the tick noise detected at 1 Hz, could not satisfy the modeling approach of a simple diffusive layer. The spectral envelopes of the S-coda waves of these events are decaying very rapidly, which suggests an origin in a more elastic environment. This is in agreement with previous studies (Giardini et al., 2020) that suggest that these events are generated deeper in the Martian mantle. For this reason, we applied another approach to these signals, with an energy envelope equation designed for deep moonquakes (Dainty et al.,1974), but it was not either capable to fit the examined data envelopes, suggesting the absence of a very thick megaregolith structure on Mars.Based on the results of the High Frequency (HF) and Very High Frequency (VF) events we observed a range of possible paths and diffusivities that can satisfy the data and we investigated the tradeoffs between the parameters of a modeling equation that control the shape of the energy envelope for the events. The analysis of these tradeoffs does not permit us to make any assumptions about the depth of the diffusive region in the Martian crust and the upper mantle as their azimuthal characteristics are unknown and therefore it is not feasible to tell if the difference in the result analysis reflects vertical or lateral variations of the uppermost diffusive layer in the Martian lithosphere.The results of this study illustrate one of the challenges in working with single-station seismic data where event location information, including distance, azimuth, and depth are crucial for understanding the lateral variation in seismic properties of a planet. The existence of a seismic network on the planetary scale will improve the ability of phase peaking and location identification of the events and therefore it will give additional constraints for a similar analysis.ReferencesClinton, J. F. et al. (2021). The Marsquake catalogue from InSight, sols 0–478.Phys Earth Planet In, 310:106595.Dainty, A. M. et al. (1974). Seismic scattering and shallow structure of the Moon in Oceanus Procellarum.The Moon,9(1-2):11–29.Giardini, D. et al. (2020). The seismicity of Mars.Nat Geosci, 13(3):205–212.InSight Marsquake Service (2021). Mars Seismic...
- Open Access
- 2 May 2024
- Published by Copernicus GmbH
The Qaanaaq airburst as an analog of seismic source in extraterrestrial atmospheres: seismic and infrasound investigation
On July 25, 2018, a meteoroid-associated airburst occurred near the Qaanaaq town, in Greenland, at approximately 22:00 UTC (20:00 local time). The event generated seismic waves that were recorded by two stations of the Danish Seismological Network (TULEG and NEEM) and the bolide trajectory was consequently calculated by the NASA Center for Near-Earth Object Studies (CNEOS). The total impact energy, calculated by CNEOS was 2.1 kT of TNT and the brightest point on its trajectory corresponds to an altitude of around 43 km, at a distance of about 50 km S of the Qaanaaq town and 50 km N of the TULEG station and the Thule Air Force Base [1].An airburst occurring over the icy surface of Greenland is a rare terrestrial analog for regions of the Solar System, where both an atmosphere and an icy surface exist. In the past, a variety of works had indicated the presence of ice on Titan, the biggest moon of Saturn (e.g. [2] and more recently [3]) and more precisely, the icy composition of mountains which are formed by tectonic activity [4]. Titan has a relatively thick atmosphere, compared to those of other moons in the Solar System, composed mainly (94%) of nitrogen [5]. The characterization of atmospheric meteoroid-associated seismic sources for Titan has a particular interest, as it is found that, contrary to other moons of the solar system, the presence of craters on its surface is extremely low (only about 0.4% according to [3]). The reason for this low cratering of the surface is the presence of the thick atmosphere, into which many of the meteoroids are entirely ablated into dust. Therefore, a methodology for the characterization of airbursts as seismic sources and the modeling of the associated generated seismic waves is necessary for a future seismic experiment, as any recorded signal will either be a direct atmospheric wave (nonlinear shock wave, or linear acoustic wave) or a seismic wave generated through the coupling of the atmospheric and solid/ice part. In the present study, our aim is to perform a seismic investigation of the Greenland ice shell with the use of the airburst-associated seismic source. The performed tasks into which this effort has been divided, include the application of a technique which approaches the bolide as an atmospheric seismic source, the calculation of the distance of shock wave propagation in the atmosphere, the description of the mechanism of generation of the seismic waves in the atmosphere and the solid-icy part.When the bolides enter the atmosphere of the Earth or that of any other body, shock waves are generated along the trajectory of the meteoroid. These waves are characterized by the overpressure that they generate, which create a clear pressure discontinuity in the atmosphere, referred to as the nonlinear part of the shock wave propagation. The propagation distance of this nonlinear wave is associated to the ratio of the meteoroid speed to the ambient sound speed, also known as the Mach number, as well as the physical diameter of the meteoroid. In this work, we compute this distance for the Earth case and for the known trajectory of the detected and examined bolide [1][6].The methodology developed in this study can serve the seismic investigation of structures covered by ice on planets or planetary bodies with a relatively thick atmosphere, where airbursts can occur due to the friction of the meteoroid with the ambient atmospheric material. An ideal example of this case are the icy mountains of Titan, which are known to be formed by tectonic activity on the Saturn’s moon [4]. The future Dragonfly mission to Titan will carry a seismometer as part of the DraGMet (Dragonfly Geophysics and Meteorology Package) payload [7]. Even if the primary goal of the mission is the characterization of the regolith properties, an eventual airburst and collection of seismic data near these mountainous icy structures, will be a great opportunity to investigate, through the identification of the associated waves and thus the investigation of the coupled seismic waves, the properties of this icy cover, its depth and composition.References: [1] https://cneos.jpl.nasa.gov/fireballs/ [2] Sohl, F. et al. (1995) Icarus, 115, 278–294 [3] Lopes R.M.C. et al. (2019) Nat Astron, [4] Radebaugh J. et al. (2007) Icarus, 192, 77-91, [5] Niemann H.B. et al. (2005) Nature, 438, 779–784 [6] Schmerr, N. et al. (2018) Abstract P21E-3406, AGU Fall Meeting 2018, Washington DC [7] Lorenz R. et al. (2018) Johns Hop- kins APL Technical Digest, 34, 3
- 2 May 2024
- Published by Copernicus GmbH
Open AccessThe Qaanaaq airburst as an analog of seismic source in extraterrestrial atmospheres: seismic and infrasound investigation
On July 25, 2018, a meteoroid-associated airburst occurred near the Qaanaaq town, in Greenland, at approximately 22:00 UTC (20:00 local time). The event generated seismic waves that were recorded by two stations of the Danish Seismological Network (TULEG and NEEM) and the bolide trajectory was consequently calculated by the NASA Center for Near-Earth Object Studies (CNEOS). The total impact energy, calculated by CNEOS was 2.1 kT of TNT and the brightest point on its trajectory corresponds to an altitude of around 43 km, at a distance of about 50 km S of the Qaanaaq town and 50 km N of the TULEG station and the Thule Air Force Base [1].An airburst occurring over the icy surface of Greenland is a rare terrestrial analog for regions of the Solar System, where both an atmosphere and an icy surface exist. In the past, a variety of works had indicated the presence of ice on Titan, the biggest moon of Saturn (e.g. [2] and more recently [3]) and more precisely, the icy composition of mountains which are formed by tectonic activity [4]. Titan has a relatively thick atmosphere, compared to those of other moons in the Solar System, composed mainly (94%) of nitrogen [5]. The characterization of atmospheric meteoroid-associated seismic sources for Titan has a particular interest, as it is found that, contrary to other moons of the solar system, the presence of craters on its surface is extremely low (only about 0.4% according to [3]). The reason for this low cratering of the surface is the presence of the thick atmosphere, into which many of the meteoroids are entirely ablated into dust. Therefore, a methodology for the characterization of airbursts as seismic sources and the modeling of the associated generated seismic waves is necessary for a future seismic experiment, as any recorded signal will either be a direct atmospheric wave (nonlinear shock wave, or linear acoustic wave) or a seismic wave generated through the coupling of the atmospheric and solid/ice part. In the present study, our aim is to perform a seismic investigation of the Greenland ice shell with the use of the airburst-associated seismic source. The performed tasks into which this effort has been divided, include the application of a technique which approaches the bolide as an atmospheric seismic source, the calculation of the distance of shock wave propagation in the atmosphere, the description of the mechanism of generation of the seismic waves in the atmosphere and the solid-icy part.When the bolides enter the atmosphere of the Earth or that of any other body, shock waves are generated along the trajectory of the meteoroid. These waves are characterized by the overpressure that they generate, which create a clear pressure discontinuity in the atmosphere, referred to as the nonlinear part of the shock wave propagation. The propagation distance of this nonlinear wave is associated to the ratio of the meteoroid speed to the ambient sound speed, also known as the Mach number, as well as the physical diameter of the meteoroid. In this work, we compute this distance for the Earth case and for the known trajectory of the detected and examined bolide [1][6].The methodology developed in this study can serve the seismic investigation of structures covered by ice on planets or planetary bodies with a relatively thick atmosphere, where airbursts can occur due to the friction of the meteoroid with the ambient atmospheric material. An ideal example of this case are the icy mountains of Titan, which are known to be formed by tectonic activity on the Saturn’s moon [4]. The future Dragonfly mission to Titan will carry a seismometer as part of the DraGMet (Dragonfly Geophysics and Meteorology Package) payload [7]. Even if the primary goal of the mission is the characterization of the regolith properties, an eventual airburst and collection of seismic data near these mountainous icy structures, will be a great opportunity to investigate, through the identification of the associated waves and thus the investigation of the coupled seismic waves, the properties of this icy cover, its depth and composition.References: [1] https://cneos.jpl.nasa.gov/fireballs/ [2] Sohl, F. et al. (1995) Icarus, 115, 278–294 [3] Lopes R.M.C. et al. (2019) Nat Astron, [4] Radebaugh J. et al. (2007) Icarus, 192, 77-91, [5] Niemann H.B. et al. (2005) Nature, 438, 779–784 [6] Schmerr, N. et al. (2018) Abstract P21E-3406, AGU Fall Meeting 2018, Washington DC [7] Lorenz R. et al. (2018) Johns Hop- kins APL Technical Digest, 34, 3
- Open Access
- 8 March 2024
- Published by Copernicus GmbH
Synthetic Seismic Waveform Analysis for Constraining Anisotropy in the D'' Layer Using the PcSdiff Phase
The D'' layer, at the base of Earth's mantle above the core-mantle boundary (CMB), is a region of profound interest due to its complex seismic characteristics and structural heterogeneity. Varying from 100 to 400 km in thickness, it features anomalous seismic velocity gradients, large-scale lateral variations, and ultra-low velocity zones, hinting at dynamic interactions with the core and potential origins for mantle plumes and subducting slabs. The PcSdiff waves, extending beyond the core-reflected PcS wave's ray theoretical limit (~63°), feature a P-wave diffracting along the CMB before being converted into an S wave. The observation of PcSdiff waves offers a novel method for examining the lowermost mantle, particularly the D'' layer, as they enable enhanced resolution in this complex region not readily available from nominal diffracted waves such as P/S diffracted waves. Their unique path can be highly sensitive to the anisotropic properties of the lowermost mantle, allowing for a detailed analysis of both isotropic and anisotropic structures, which help elucidate the dynamics and composition of this enigmatic region. This study focuses on an in-depth exploration of the D'' layer's anisotropic properties, and performs seismic forward modeling against the novel PcSdiff phase. By simulating seismic wave propagation through this region, we aim to provide a novel methodological approach to decode the intricate seismic interactions within the D'' layer. In the first step, we use an extensive synthetic waveform dataset to understand the responses of the PcSdiff phase to the geometry of the D'' layer and its anisotropic features. This process is pivotal in isolating and understanding the PcSdiff phase's nuanced responses to radial and lateral velocity variations due to anisotropy within the D'' layer. We aim to dissect these responses in systematic detail, looking for subtle signatures and patterns that might be indicative of specific anisotropic features. Then, our approach integrates seismic data from the USArray with synthetics produced for 3D models of the internal structure. In pursuing this detailed characterization, our objective is to provide a more nuanced and comprehensive understanding of how the novel PcSdiff wave can be used to probe the anisotropic properties of the D'' layer.
- 8 March 2024
- Published by Copernicus GmbH
Open AccessSynthetic Seismic Waveform Analysis for Constraining Anisotropy in the D'' Layer Using the PcSdiff Phase
The D'' layer, at the base of Earth's mantle above the core-mantle boundary (CMB), is a region of profound interest due to its complex seismic characteristics and structural heterogeneity. Varying from 100 to 400 km in thickness, it features anomalous seismic velocity gradients, large-scale lateral variations, and ultra-low velocity zones, hinting at dynamic interactions with the core and potential origins for mantle plumes and subducting slabs. The PcSdiff waves, extending beyond the core-reflected PcS wave's ray theoretical limit (~63°), feature a P-wave diffracting along the CMB before being converted into an S wave. The observation of PcSdiff waves offers a novel method for examining the lowermost mantle, particularly the D'' layer, as they enable enhanced resolution in this complex region not readily available from nominal diffracted waves such as P/S diffracted waves. Their unique path can be highly sensitive to the anisotropic properties of the lowermost mantle, allowing for a detailed analysis of both isotropic and anisotropic structures, which help elucidate the dynamics and composition of this enigmatic region. This study focuses on an in-depth exploration of the D'' layer's anisotropic properties, and performs seismic forward modeling against the novel PcSdiff phase. By simulating seismic wave propagation through this region, we aim to provide a novel methodological approach to decode the intricate seismic interactions within the D'' layer. In the first step, we use an extensive synthetic waveform dataset to understand the responses of the PcSdiff phase to the geometry of the D'' layer and its anisotropic features. This process is pivotal in isolating and understanding the PcSdiff phase's nuanced responses to radial and lateral velocity variations due to anisotropy within the D'' layer. We aim to dissect these responses in systematic detail, looking for subtle signatures and patterns that might be indicative of specific anisotropic features. Then, our approach integrates seismic data from the USArray with synthetics produced for 3D models of the internal structure. In pursuing this detailed characterization, our objective is to provide a more nuanced and comprehensive understanding of how the novel PcSdiff wave can be used to probe the anisotropic properties of the D'' layer.
- Open Access
- 3 September 2023
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Planets
- Vol. 128 (9)
Focal Mechanism Determination of Event S1222a and Implications for Tectonics Near the Dichotomy Boundary in Southern Elysium Planitia, Mars
On 4 May 2022 the InSight seismometer SEIS‐VBB recorded the largest marsquake ever observed, S1222a, with an initial magnitude estimate of 4.6. Understanding the depth and source properties of this event has important implications for the nature of tectonic activity on Mars. Located ∼37° to the southeast of InSight, S1222a is one of the few non‐impact marsquakes that exhibits prominent surface waves. We use waveform modeling of body waves (P and S) and surface waves (Rayleigh and Love) to constrain the focal mechanism, assuming a double‐couple source, and find that S1222a likely resulted from reverse faulting in the crust (source depth near 22 km). We estimate the scalar moment to be 2.5 × 1015–3.5 × 1015 Nm (magnitude MW 4.2–4.3). Our results suggest active compressional tectonics near the dichotomy boundary on Mars, likely due to thermal contraction from planetary cooling.
- 3 September 2023
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research: Planets
- Vol. 128 (9)
Open AccessFocal Mechanism Determination of Event S1222a and Implications for Tectonics Near the Dichotomy Boundary in Southern Elysium Planitia, Mars
On 4 May 2022 the InSight seismometer SEIS‐VBB recorded the largest marsquake ever observed, S1222a, with an initial magnitude estimate of 4.6. Understanding the depth and source properties of this event has important implications for the nature of tectonic activity on Mars. Located ∼37° to the southeast of InSight, S1222a is one of the few non‐impact marsquakes that exhibits prominent surface waves. We use waveform modeling of body waves (P and S) and surface waves (Rayleigh and Love) to constrain the focal mechanism, assuming a double‐couple source, and find that S1222a likely resulted from reverse faulting in the crust (source depth near 22 km). We estimate the scalar moment to be 2.5 × 1015–3.5 × 1015 Nm (magnitude MW 4.2–4.3). Our results suggest active compressional tectonics near the dichotomy boundary on Mars, likely due to thermal contraction from planetary cooling.
- Open Access
- 27 July 2023
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 235 (2) , 1725-1735
Reproducing complex anisotropy patterns at subduction zones from splitting intensity analysis and anisotropy tomography
Measurements of seismic anisotropy provide a lot of information on the deformation and structure as well as flows of the Earth's interior, in particular of the upper mantle. Even though the strong and heterogeneous seismic anisotropic nature of the upper mantle has been demonstrated by a wealth of theoretical and observational approaches , most of standard teleseismic body-wave tomography studies overlook P- and S-wave anisotropy, thus producing artefacts in tomographic models in terms of amplitude and localization of heterogeneities. Conventional methods of seismic anisotropy measurement have their limitations regarding lateral and mainly depth resolution. To overcome this problem much effort has been done to develop tomographic methods to invert shear wave splitting data for anisotropic structures, based on finite-frequency sensitivity kernels that relate model perturbations to splitting observations. A promising approach to image the upper mantle anisotropy is the inversion of splitting intensity (SI). This seismic observable is a measure of the amount of energy on the transverse component waveform and, to a first order, it is linearly related to the elastic perturbations of the medium through the 3-D sensitivity kernels, that can be therefore inverted, allowing a high-resolution image of the upper mantle anisotropy. Here we present an application of the SI tomography to a synthetic subduction setting. Starting from synthetic SKS waveforms, we first derived high-quality SKS SI measurements; then we used the SI data as input into tomographic inversion. This approach enables high‐resolution tomographic images of upper‐mantle anisotropy through recovering vertical and lateral changes in anisotropy and represents a propaedeutic step to the real cases of subduction settings. Additionally this study was able to detect regions of strong dipping anisotropy by allowing a 360° periodic dependence of the splitting vector.
- 27 July 2023
- Published by Oxford University Press (OUP) in Geophysical Journal International
- Vol. 235 (2) , 1725-1735
Open AccessReproducing complex anisotropy patterns at subduction zones from splitting intensity analysis and anisotropy tomography
Measurements of seismic anisotropy provide a lot of information on the deformation and structure as well as flows of the Earth's interior, in particular of the upper mantle. Even though the strong and heterogeneous seismic anisotropic nature of the upper mantle has been demonstrated by a wealth of theoretical and observational approaches , most of standard teleseismic body-wave tomography studies overlook P- and S-wave anisotropy, thus producing artefacts in tomographic models in terms of amplitude and localization of heterogeneities. Conventional methods of seismic anisotropy measurement have their limitations regarding lateral and mainly depth resolution. To overcome this problem much effort has been done to develop tomographic methods to invert shear wave splitting data for anisotropic structures, based on finite-frequency sensitivity kernels that relate model perturbations to splitting observations. A promising approach to image the upper mantle anisotropy is the inversion of splitting intensity (SI). This seismic observable is a measure of the amount of energy on the transverse component waveform and, to a first order, it is linearly related to the elastic perturbations of the medium through the 3-D sensitivity kernels, that can be therefore inverted, allowing a high-resolution image of the upper mantle anisotropy. Here we present an application of the SI tomography to a synthetic subduction setting. Starting from synthetic SKS waveforms, we first derived high-quality SKS SI measurements; then we used the SI data as input into tomographic inversion. This approach enables high‐resolution tomographic images of upper‐mantle anisotropy through recovering vertical and lateral changes in anisotropy and represents a propaedeutic step to the real cases of subduction settings. Additionally this study was able to detect regions of strong dipping anisotropy by allowing a 360° periodic dependence of the splitting vector.
- Open Access
- 15 May 2023
- Published by Copernicus GmbH
SPHY3D: A hybrid seismic computational framework for box-tomography of spherical Earth
Computational seismology encountered a dramatic advance during the past decades with the development of SEM codes that use the simultaneous increase of the available computational power. Meanwhile, the use of teleseismic events for regional seismic tomography is suggested with the application of the box-tomography methodology (Masson and Romanowicz, 2017). In this work we use these advances in order to suggest a package for box-tomography, using AxiSEM for 1-D global wavefield simulations (Nissen-Meyer et al., 2014) and SPECFEM3D for 3-D regional seismic simulations (Komatitsch and Tromp, 1999). These codes have been previously used and validated for such hybrid simulations (Monteiller et al., 2021), however with the limitation on the dimensions of the examined region, where 3-D full waveform topography is applied, due to the Cartesian setting that does not honour the curvature of the Earth. Although recent advances solved this limitation for SPECFEM3D Global, by permitting the use of a small Earth chunk, the Cartesian description of the regional model allows computing the injection of the 1-D computed wavefield from the global model to the regional box. Therefore, we developed and present comparative results of a package that transforms the geometry of the Cartesian simulation in a "spherical Earth" setting and allows the performance of hybrid simulations for box tomography in regions larger than a couple of degrees. The code changes the shape of a Cartesian rectangular mesh into a curved one and through a series of interpolations adjusts the geometry of any given structure model, the topography of the surface and the interfaces, and the position of the receivers. The simulations are tested against real data, as we perform our computations on a dynamically interesting area, with the presence of a subduction slab in the central Mediterranean. We test the methodology on seismological inverse models for the local structure (Rappisi et al., 2021).References:Komatitsch, D. and Tromp, J., 1999. Introduction to the spectral element method for three-dimensional seismic wave propagation. Geophysical journal international, 139(3), pp.806-822.Masson, Y. and Romanowicz, B., 2017. Box tomography: localized imaging of remote targets buried in an unknown medium, a step forward for understanding key structures in the deep Earth. Geophysical Journal International, 211(1), pp.141-163.Monteiller, V., Beller, S., Plazolles, B. and Chevrot, S., 2021. On the validity of the planar wave approximation to compute synthetic seismograms of teleseismic body waves in a 3-D regional model. Geophysical Journal International, 224(3), pp.2060-2076.Nissen-Meyer, T., van Driel, M., Stähler, S.C., Hosseini, K., Hempel, S., Auer, L., Colombi, A. and Fournier, A., 2014. AxiSEM: broadband 3-D seismic wavefields in axisymmetric media. Solid Earth, 5(1), pp.425-445.Rappisi, F., VanderBeek, B.P., Faccenda, M., Morelli, A. and Molinari, I., 2022. Slab geometry and upper mantle flow patterns in the Central Mediterranean from 3D anisotropic P-wave tomography. Journal of Geophysical Research: Solid Earth, p.e2021JB023488.
- 15 May 2023
- Published by Copernicus GmbH
Open AccessSPHY3D: A hybrid seismic computational framework for box-tomography of spherical Earth
Computational seismology encountered a dramatic advance during the past decades with the development of SEM codes that use the simultaneous increase of the available computational power. Meanwhile, the use of teleseismic events for regional seismic tomography is suggested with the application of the box-tomography methodology (Masson and Romanowicz, 2017). In this work we use these advances in order to suggest a package for box-tomography, using AxiSEM for 1-D global wavefield simulations (Nissen-Meyer et al., 2014) and SPECFEM3D for 3-D regional seismic simulations (Komatitsch and Tromp, 1999). These codes have been previously used and validated for such hybrid simulations (Monteiller et al., 2021), however with the limitation on the dimensions of the examined region, where 3-D full waveform topography is applied, due to the Cartesian setting that does not honour the curvature of the Earth. Although recent advances solved this limitation for SPECFEM3D Global, by permitting the use of a small Earth chunk, the Cartesian description of the regional model allows computing the injection of the 1-D computed wavefield from the global model to the regional box. Therefore, we developed and present comparative results of a package that transforms the geometry of the Cartesian simulation in a "spherical Earth" setting and allows the performance of hybrid simulations for box tomography in regions larger than a couple of degrees. The code changes the shape of a Cartesian rectangular mesh into a curved one and through a series of interpolations adjusts the geometry of any given structure model, the topography of the surface and the interfaces, and the position of the receivers. The simulations are tested against real data, as we perform our computations on a dynamically interesting area, with the presence of a subduction slab in the central Mediterranean. We test the methodology on seismological inverse models for the local structure (Rappisi et al., 2021).References:Komatitsch, D. and Tromp, J., 1999. Introduction to the spectral element method for three-dimensional seismic wave propagation. Geophysical journal international, 139(3), pp.806-822.Masson, Y. and Romanowicz, B., 2017. Box tomography: localized imaging of remote targets buried in an unknown medium, a step forward for understanding key structures in the deep Earth. Geophysical Journal International, 211(1), pp.141-163.Monteiller, V., Beller, S., Plazolles, B. and Chevrot, S., 2021. On the validity of the planar wave approximation to compute synthetic seismograms of teleseismic body waves in a 3-D regional model. Geophysical Journal International, 224(3), pp.2060-2076.Nissen-Meyer, T., van Driel, M., Stähler, S.C., Hosseini, K., Hempel, S., Auer, L., Colombi, A. and Fournier, A., 2014. AxiSEM: broadband 3-D seismic wavefields in axisymmetric media. Solid Earth, 5(1), pp.425-445.Rappisi, F., VanderBeek, B.P., Faccenda, M., Morelli, A. and Molinari, I., 2022. Slab geometry and upper mantle flow patterns in the Central Mediterranean from 3D anisotropic P-wave tomography. Journal of Geophysical Research: Solid Earth, p.e2021JB023488.
- Open Access
- 15 May 2023
- Published by Copernicus GmbH
Testing the splitting intensity methodology to retrieve average, dipping, and depth dependent anisotropy from a complex subduction model
Seismic anisotropy measurements provide a lot of information on the deformation and structure of the Earth’s interior, in particular of the upper mantle. Conventional methods of measurement of anisotropy have their limitations, especially regarding depth resolution. Splitting intensity (SI) is a seismic observable, related to the amount of energy on the transverse component waveform and, to a first order, it is linearly related to the elastic perturbations of the medium through the 3-D sensitivity kernels, that can be therefore inverted, allowing a high-resolution image of the upper-mantle anisotropy. Starting from synthetic SKS waveforms, we first derived high-quality SKS splitting intensity measurements; then we used the splitting intensity data as input into tomographic inversion. This approach enables high‐resolution tomographic images of horizontal upper‐mantle anisotropy through recovering vertical and lateral changes in anisotropy and represents a propaedeutic step to the real cases of subduction settings. Additionally, this approach was able to detect regions of strong dipping anisotropy by allowing a 360° periodic dependence of the splitting vector. Single and thick layers of dipping angles between 30 and 60° are clearly represented with a high dt2 value, while double layers or nearly vertical dips are more difficult to identify.
- 15 May 2023
- Published by Copernicus GmbH
Open AccessTesting the splitting intensity methodology to retrieve average, dipping, and depth dependent anisotropy from a complex subduction model
Seismic anisotropy measurements provide a lot of information on the deformation and structure of the Earth’s interior, in particular of the upper mantle. Conventional methods of measurement of anisotropy have their limitations, especially regarding depth resolution. Splitting intensity (SI) is a seismic observable, related to the amount of energy on the transverse component waveform and, to a first order, it is linearly related to the elastic perturbations of the medium through the 3-D sensitivity kernels, that can be therefore inverted, allowing a high-resolution image of the upper-mantle anisotropy. Starting from synthetic SKS waveforms, we first derived high-quality SKS splitting intensity measurements; then we used the splitting intensity data as input into tomographic inversion. This approach enables high‐resolution tomographic images of horizontal upper‐mantle anisotropy through recovering vertical and lateral changes in anisotropy and represents a propaedeutic step to the real cases of subduction settings. Additionally, this approach was able to detect regions of strong dipping anisotropy by allowing a 360° periodic dependence of the splitting vector. Single and thick layers of dipping angles between 30 and 60° are clearly represented with a high dt2 value, while double layers or nearly vertical dips are more difficult to identify.
- Open Access
- 1 March 2023
- Published by Wiley
Moment Tensor Estimation of Event S1222a and Implications for Tectonics Near the Dichotomy Boundary in Southern Elysium Planitia, Mars
On May 4th, 2022 the InSight seismometer SEIS recorded the largest marsquake ever observed, S1222a, with an initial magnitude estimate of Mw 4.7. Understanding the depth and source properties of this event has important implications for the nature of tectonic activity on Mars. Located ~37 degrees to the southeast of InSight, S1222a is one of the few non-impact marsquakes that exhibits prominent ratio surface waves. We use waveform modeling of body waves (P and S) and surface waves (Rayleigh and Love) to constrain the moment tensor and quantify the associated uncertainty. We find that S1222a likely resulted from dip-slip faulting in the mid-crust (source depth ~18 – 28 km) and estimate a scalar moment of 3.51015 – 5.01015 Nm (magnitude Mw 4.3 – 4.4). The best-fitting focal mechanism is sensitive to the choice of phase windows and misfit weights, as well as the structural model of Mars used to calculate Green’s functions. We find that an E-W to SE-NW striking thrust fault can explain the data well, although depending on the choice of misfit weighting, a normal fault solution is also permissible. The orientation of the best-fitting fault plane solutions suggests that S1222a takes place on a fault system near the martian crustal dichotomy accommodating relative motion between the northern lowlands and southern highlands. Independent constraints on the event depth and improved models of the (an)isotropic velocity structure of the martian crust and mantle could help resolve the ambiguity inherent to single-station moment tensor inversions of S1222a and other marsquakes.
- 1 March 2023
- Published by Wiley
Open AccessMoment Tensor Estimation of Event S1222a and Implications for Tectonics Near the Dichotomy Boundary in Southern Elysium Planitia, Mars
On May 4th, 2022 the InSight seismometer SEIS recorded the largest marsquake ever observed, S1222a, with an initial magnitude estimate of Mw 4.7. Understanding the depth and source properties of this event has important implications for the nature of tectonic activity on Mars. Located ~37 degrees to the southeast of InSight, S1222a is one of the few non-impact marsquakes that exhibits prominent ratio surface waves. We use waveform modeling of body waves (P and S) and surface waves (Rayleigh and Love) to constrain the moment tensor and quantify the associated uncertainty. We find that S1222a likely resulted from dip-slip faulting in the mid-crust (source depth ~18 – 28 km) and estimate a scalar moment of 3.51015 – 5.01015 Nm (magnitude Mw 4.3 – 4.4). The best-fitting focal mechanism is sensitive to the choice of phase windows and misfit weights, as well as the structural model of Mars used to calculate Green’s functions. We find that an E-W to SE-NW striking thrust fault can explain the data well, although depending on the choice of misfit weighting, a normal fault solution is also permissible. The orientation of the best-fitting fault plane solutions suggests that S1222a takes place on a fault system near the martian crustal dichotomy accommodating relative motion between the northern lowlands and southern highlands. Independent constraints on the event depth and improved models of the (an)isotropic velocity structure of the martian crust and mantle could help resolve the ambiguity inherent to single-station moment tensor inversions of S1222a and other marsquakes.
- Open Access
- 10 October 2022
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 119 (42)
Seismic detection of a deep mantle discontinuity within Mars by InSight
Constraining the thermal and compositional state of the mantle is crucial for deciphering the formation and evolution of Mars. Mineral physics predicts that Mars’ deep mantle is demarcated by a seismic discontinuity arising from the pressure-induced phase transformation of the mineral olivine to its higher-pressure polymorphs, making the depth of this boundary sensitive to both mantle temperature and composition. Here, we report on the seismic detection of a midmantle discontinuity using the data collected by NASA’s InSight Mission to Mars that matches the expected depth and sharpness of the postolivine transition. In five teleseismic events, we observed triplicated P and S waves and constrained the depth of this discontinuity to be 1,006 ± 40 km by modeling the triplicated waveforms. From this depth range, we infer a mantle potential temperature of 1,605 ± 100 K, a result consistent with a crust that is 10 to 15 times more enriched in heat-producing elements than the underlying mantle. Our waveform fits to the data indicate a broad gradient across the boundary, implying that the Martian mantle is more enriched in iron compared to Earth. Through modeling of thermochemical evolution of Mars, we observe that only two out of the five proposed composition models are compatible with the observed boundary depth. Our geodynamic simulations suggest that the Martian mantle was relatively cold 4.5 Gyr ago (1,720 to 1,860 K) and are consistent with a present-day surface heat flow of 21 to 24 mW/m 2 .
- 10 October 2022
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 119 (42)
Open AccessSeismic detection of a deep mantle discontinuity within Mars by InSight
Constraining the thermal and compositional state of the mantle is crucial for deciphering the formation and evolution of Mars. Mineral physics predicts that Mars’ deep mantle is demarcated by a seismic discontinuity arising from the pressure-induced phase transformation of the mineral olivine to its higher-pressure polymorphs, making the depth of this boundary sensitive to both mantle temperature and composition. Here, we report on the seismic detection of a midmantle discontinuity using the data collected by NASA’s InSight Mission to Mars that matches the expected depth and sharpness of the postolivine transition. In five teleseismic events, we observed triplicated P and S waves and constrained the depth of this discontinuity to be 1,006 ± 40 km by modeling the triplicated waveforms. From this depth range, we infer a mantle potential temperature of 1,605 ± 100 K, a result consistent with a crust that is 10 to 15 times more enriched in heat-producing elements than the underlying mantle. Our waveform fits to the data indicate a broad gradient across the boundary, implying that the Martian mantle is more enriched in iron compared to Earth. Through modeling of thermochemical evolution of Mars, we observe that only two out of the five proposed composition models are compatible with the observed boundary depth. Our geodynamic simulations suggest that the Martian mantle was relatively cold 4.5 Gyr ago (1,720 to 1,860 K) and are consistent with a present-day surface heat flow of 21 to 24 mW/m 2 .