E. Baron

University of Oklahoma, Norman, Oklahoma, United States

Are you E. Baron?

Claim your profile

Publications (276)898.61 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: We present the intensive spectroscopic follow up of the Type Ia supernova (SN Ia) 2014J in the starburst galaxy M82. Twenty-seven optical spectra have been acquired from 2014 January 22 to September 1 with the Isaac Newton and William Herschel Telescopes. After correcting the observations for the recession velocity of M82 and for Milky Way and host galaxy extinction, we measured expansion velocities from spectral line blueshifts and pseudo-equivalent width of the strongest features in the spectra, which gives an idea on how elements are distributed within the ejecta. We position SN 2014J in the Benetti, Branch et al. and Wang et al. diagrams. These diagrams are based on properties of the Si ii features and provide dynamical and chemical information about the SN ejecta. The nearby SN 2011fe, which showed little evidence for reddening in its host galaxy, is shown as a reference for comparisons. SN 2014J is a border-line object between the Core-normal and Broad-line groups, which corresponds to an intermediate position between low-velocity gradient and high-velocity gradient objects. SN 2014J follows the $R({\rm Si\,{\small {II}}})$–Δm15 correlation, which confirms its classification as a relatively normal SN Ia. Our description of the SN Ia in terms of the evolution of the pseudo-equivalent width of various ions as well as the position in the various diagrams put this specific SN Ia into the overall sample of SN Ia.
    No preview · Article · Mar 2016 · Monthly Notices of the Royal Astronomical Society
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We use observed UV through near IR spectra to examine whether SN 2011fe can be understood in the framework of Branch-normal SNe Ia and to examine its individual peculiarities. As a benchmark, we use a delayed-detonation model with a progenitor metallicity of Z_solar/20. We study the sensitivity of features to variations in progenitor metallicity, the outer density profile, and the distribution of radioactive nickel. The effect of metallicity variations in the progenitor have a relatively small effect on the synthetic spectra. We also find that the abundance stratification of SN 2011fe resembles closely that of a delayed detonation model with a transition density that has been fit to other Branch-normal Type Ia supernovae. At early times, the model photosphere is formed in material with velocities that are too high, indicating that the photosphere recedes too slowly or that SN 2011fe has a lower specific energy in the outer ~0.1 M_sun than does the model. We discuss several explanations for the discrepancies. Finally, we examine variations in both the spectral energy distribution and in the colors due to variations in the progenitor metallicity, which suggests that colors are only weak indicators for the progenitor metallicity, in the particular explosion model that we have studied. We do find that the flux in the U band is significantly higher at maximum light in the solar metallicity model than in the lower metallicity model and the lower metallicity model much better matches the observed spectrum.
    Full-text · Article · Aug 2015 · Monthly Notices of the Royal Astronomical Society
  • [Show abstract] [Hide abstract]
    ABSTRACT: Active galactic nuclei (AGNs) and quasars are important astrophysical objects to understand. Recently, microlensing observations have constrained the size of the quasar X-ray emission region to be of the order of 10 gravitational radii of the central supermassive black hole. For distances within a few gravitational radii, light paths are strongly bent by the strong gravity field of the central black hole. If the central black hole has nonzero angular momentum (spin), a photon's polarization plane will be rotated by the gravitational Faraday effect. The observed X-ray flux and polarization will then be influenced significantly by the strong gravity field near the source. Consequently, linear gravitational lensing theory is inadequate for such extreme circumstances. We present simple algorithms computing strong lensing effects of Kerr black holes, including effects on polarization. Our algorithms are realized in a program "KERTAP" in two versions: MATLAB and Python. The key ingredients of KERTAP are: a graphic user interface, a {\it backward} ray-tracing algorithm, a polarization propagator dealing with gravitational Faraday rotation, and algorithms computing observables such as flux magnification and polarization angles. Our algorithms can be easily realized in other programming languages such as FORTRAN, C, and C++. The MATLAB version of KERTAP is parallelized using the MATLAB Parallel Computing Toolbox and the Distributed Computing Server. The Python code was sped up using Cython and supports full implementation of MPI using 'mpi4py' package. As an example, we investigate the inclination angle dependence of the observed polarization and the strong lensing magnification of AGN X-ray emission. We conclude that it is possible to perform complex numerical-relativity-related computations using interpreted languages such as MATLAB and Python.
    No preview · Article · May 2015 · The Astrophysical Journal Supplement Series
  • [Show abstract] [Hide abstract]
    ABSTRACT: The effect of metallicity on the observed light of Type Ia supernovae (SNe Ia) could lead to systematic errors as the absolute magnitudes of local and distant SNe Ia are compared to measure luminosity distances and determine cosmological parameters. The UV light may be especially sensitive to metallicity, though different modeling methods disagree as to the magnitude, wavelength dependence, and even the sign of the effect. The outer density structure, ^56 Ni, and to a lesser degree asphericity, also impact the UV. We compute synthetic photometry of various metallicity-dependent models and compare to UV/optical photometry from the Swift Ultra-Violet/Optical Telescope. We find that the scatter in the mid-UV to near-UV colors is larger than predicted by changes in metallicity alone and is not consistent with reddening. We demonstrate that a recently employed method to determine relative abundances using UV spectra can be done using UVOT photometry, but we warn that accurate results require an accurate model of the cause of the variations. The abundance of UV photometry now available should provide constraints on models that typically rely on UV spectroscopy for constraining metallicity, density, and other parameters. Nevertheless, UV spectroscopy for a variety of SN explosions is still needed to guide the creation of accurate models. A better understanding of the influences affecting the UV is important for using SNe Ia as cosmological probes, as the UV light may test whether SNe Ia are significantly affected by evolutionary effects.
    No preview · Article · Apr 2015 · The Astrophysical Journal
  • Source
    D. Jack · E. Baron · P. H. Hauschildt
    [Show abstract] [Hide abstract]
    ABSTRACT: We obtained a time series of spectra covering the secondary maximum in the I band of the bright Type Ia supernova 2014J in M82 with the TIGRE telescope. Comparing the observations with theoretical models calculated with the time dependent extension of the phoenix code, we identify the feature that causes the secondary maximum in the I-band light curve. Fe ii 3d6(3D)4s–3d6(5D)4p and similar high-excitation transitions produce a blended feature at ∼7500 Å, which causes the rise of the light curve towards the secondary maximum. The series of observed spectra of SN 2014J and archival data of SN 2011fe confirm this conclusion. We further studied the plateau phase of the R-band light curve of SN 2014J and searched for features which contribute to the flux. The theoretical models do not clearly indicate a new feature that may cause the R-band plateau phase. However, Co ii features in the range of 6500–7000 Å and the Fe ii feature of the I band are clearly seen in the theoretical spectra, but do not appear to provide all of the flux necessary for the R-band plateau.
    Preview · Article · Mar 2015 · Monthly Notices of the Royal Astronomical Society
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present UV through NIR broad-band photometry, and optical and NIR spectroscopy of Type Iax supernova 2012Z. The data set consists of both early and late-time observations, including the first late phase NIR spectrum obtained for a spectroscopically classified SN Iax. Simple model calculations of its bolometric light curve suggest SN 2012Z produced ~0.3 M_sun of (56)Ni, ejected about a Chandrasekhar mass of material, and had an explosion energy of ~10^51 erg, making it one of the brightest and most energetic SN Iax yet observed. The late phase NIR spectrum of SN 2012Z is found to broadly resemble similar epoch spectra of normal SNe Ia; however, like other SNe Iax, corresponding visual-wavelength spectra differ substantially compared to all supernova types. Constraints from the distribution of IMEs, e.g. silicon and magnesium, indicate that the outer ejecta did not experience significant mixing during or after burning, and the late phase NIR line profiles suggests most of the (56)Ni is produced during high density burning. The various observational properties of SN 2012Z are found to be consistent with the theoretical expectations of a Chandrasekhar mass white dwarf progenitor that experiences a pulsational delayed detonation, which produced several tenths of a solar mass of (56)Ni during the deflagration burning phase and little (or no) (56)Ni during the detonation phase. Within this scenario only a moderate amount of Rayleigh-Taylor mixing occurs both during the deflagration and fallback phase of the pulsation, and the layered structure of the IMEs is a product of the subsequent denotation phase. The fact that the SNe Iax population does not follow a tight brightness-decline relation similar to SNe Ia can then be understood in the framework of variable amounts of mixing during pulsational rebound and variable amounts of (56)Ni production during the early subsonic phase of expansion.
    Full-text · Article · Aug 2014 · Astronomy and Astrophysics
  • Source
    E. Baron
    [Show abstract] [Hide abstract]
    ABSTRACT: The question of the nature of the progenitor of Type Ia supernovae (SNe Ia) is important both for our detailed understanding of stellar evolution and for their use as cosmological probes of the dark energy. Much of the basic features of SNe Ia can be understood directly from the nuclear physics, a fact which Gerry would have appreciated. We present an overview of the current observational and theoretical situation and show that it not incompatible with most SNe Ia being the results of thermonuclear explosions near the Chandrasekhar mass.
    Preview · Article · Aug 2014 · Nuclear Physics A
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present near-infrared synthetic spectra of a delayed-detonation hydrodynamical model and compare them to observed spectra of four normal type Ia supernovae ranging from day +56.5 to day +85. This is the epoch during which supernovae are believed to be undergoing the transition from the photospheric phase, where spectra are characterized by line scattering above an optically thick photosphere, to the nebular phase, where spectra consist of optically thin emission from forbidden lines. We find that most spectral features in the near-infrared can be accounted for by permitted lines of Fe II and Co II. In addition, we find that [Ni II] fits the emission feature near 1.98 {\mu}m, suggesting that a substantial mass of 58Ni exists near the center of the ejecta in these objects, arising from nuclear burning at high density. A tentative identification of Mn II at 1.15 {\mu}m may support this conclusion as well.
    Preview · Article · Jul 2014 · The Astrophysical Journal
  • [Show abstract] [Hide abstract]
    ABSTRACT: We present the first J-band spectrum of Mrk 231, which reveals a large He I* λ10830 broad absorption line with a profile similar to that of the well-known Na I broad absorption line. Combining this spectrum with optical and UV spectra from the literature, we show that the unusual reddening noted by Veilleux et al. is explained by a reddening curve like those previously used to explain low values of total-to-selective extinction in Type Ia supernovae. The nuclear starburst may be the origin and location of the dust. Spatially resolved emission in the broad absorption line trough suggests nearly full coverage of the continuum emission region. The broad absorption lines reveal higher velocities in the He I* lines (produced in the quasar-photoionized H II region) compared with the Na I and Ca II lines (produced in the corresponding partially ionized zone). Cloudy simulations show that a density increase is required between the H II and partially ionized zones to produce ionic column densities consistent with the optical and IR absorption line measurements and limits, and that the absorber lies ~100 pc from the central engine. These results suggest that the He I* lines are produced in an ordinary quasar BAL wind that impacts upon, compresses, and accelerates the nuclear starburst's dusty effluent (feedback in action), and the Ca II and Na I lines are produced in this dusty accelerated gas. This unusual circumstance explains the rarity of Na I absorption lines; without the compression along our line of sight, Mrk 231 would appear as an ordinary iron low-ionization, broad absorption line quasar.
    No preview · Article · May 2014 · The Astrophysical Journal
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present the first J-band spectrum of Mrk 231, which reveals a large \ion{He}{1}*$\lambda 10830$ broad absorption line with a profile similar to that of the well-known \ion{Na}{1} broad absorption line. Combining this spectrum with optical and UV spectra from the literature, we show that the unusual reddening noted by \citet{veilleux13} is explained by a reddening curve like those previously used to explain low values of total-to-selective extinction in SNe Ia. The nuclear starburst may be the origin and location of the dust. Spatially-resolved emission in the broad absorption line trough suggests nearly full coverage of the continuum emission region. The broad absorption lines reveal higher velocities in the \ion{He}{1}* lines (produced in the quasar-photoionized \ion{H}{2} region) compared with the \ion{Na}{1} and \ion{Ca}{2} lines (produced in the corresponding partially-ionized zone). {\it Cloudy} simulations show that a density increase is required between the \ion{H}{2} and partially-ionized zones to produce ionic column densities consistent with the optical and IR absorption line measurements and limits, and that the absorber lies $\sim 100\rm \, pc$ from the central engine. These results suggest that the \ion{He}{1}* lines are produced in an ordinary quasar BAL wind that impacts upon, compresses, and accelerates the nuclear starburst's dusty effluent (feedback in action), and the \ion{Ca}{2} and \ion{Na}{1} lines are produced in this dusty accelerated gas. This unusual circumstance explains the rarity of \ion{Na}{1} absorption lines; without the compression along our line of sight, Mrk~231 would appear as an ordinary FeLoBAL.
    Preview · Article · Apr 2014

  • No preview · Article · Mar 2014
  • [Show abstract] [Hide abstract]
    ABSTRACT: With substantial target-of-opportunity (ToO) access on Gemini- South (+GMOS) and the VLT (+X-Shooter), along with visitor nights at the NTT (+EFOSC, SOFI), SOAR (+GOODMAN), and du Pont (+WFCCD) telescopes, a detailed time series of optical and NIR spectroscopy was obtained for SN 2010ae. The resulting early phase time series consists of 21 spectra covering 20 epochs of optical spectroscopy, extending from -2d to +57d relative to T(B)max, as well as eight NIR spectra covering seven epochs ranging from -1d to +18d. Additionally, at late phases a visual-wavelength spectrum was taken with the VLT (+FORS2) on +252d.(2 data files).
    No preview · Article · Dec 2013
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: A comprehensive set of optical and near-infrared (NIR) photometry and spectroscopy is presented for the faint and fast 2008ha-like supernova (SN) 2010ae. Contingent on the adopted value of host extinction SN 2010ae reached a peak brightness of ~ -13.8 > M_V > -15.3 mag, while modeling of the UVOIR light curve suggests it produced 0.003--0.007 M_sun of 56^Ni, ejected ~ 0.30--0.60 M_sun of material, and had an explosion energy of ~ 0.04--0.30x10^51 erg. The values of these explosion parameters are similar to the peculiar SN 2008ha --for which we also present previously unpublished early phase optical and NIR light curves-- and places these two transients at the faint end of the 2002cx-like SN population. Detailed inspection of the post maximum NIR spectroscopic sequence indicates the presence of a multitude of spectral features, which are identified through SYNAPPS modeling to be mainly attributed to Co II. Comparison with a collection of published and unpublished NIR spectra of other 2002cx-like SNe, reveals that a Co II footprint is ubiquitous to this subclass of transients, providing a link to Type Ia SNe. A visual-wavelength spectrum of SN 2010ae obtained +252 days past maximum shows a striking resemblance to a similar epoch spectrum of SN 2002cx. However subtle difference in the strength and ratio of calcium emission features, as well as diversity amongst similar epoch spectra of other 2002cx-like SNe indicates a range of physical conditions of the ejecta, highlighting the heterogeneous nature of this peculiar class of transients.
    Full-text · Article · Nov 2013 · Astronomy and Astrophysics
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We present predictions for hydrogen and helium emission line luminosities from circumstellar matter around Type Ia supernovae (SNe Ia) using time dependent photoionization modeling. ESO/VLT optical echelle spectra of the SN Ia 2000cx were taken before and up to 70 days after maximum. We detect no hydrogen and helium lines, and place an upper limit on the mass loss rate for the putative wind of less than 1.3EE{-5} solar masses per year, assuming a speed of 10 km/s and solar abundances for the wind. In a helium-enriched case, the best line to constrain the mass loss would be He I 10,830 A. We confirm the details of interstellar Na I and Ca II absorption towards SN 2000cx as discussed by Patat et al., but also find evidence for 6613.56 A Diffuse Interstellar Band (DIB) absorption in the Milky Way. We discuss measurements of the X-ray emission from the interaction between the supernova ejecta and the wind and we re-evaluate observations of SN 1992A obtained 16 days after maximum by Schlegel & Petre. We find an upper limit of 1.3EE{-5} solar masses per year. These results, together with the previous observational work on the normal SNe Ia 1994D and 2001el, disfavour a symbiotic star in the upper mass loss rate regime from being the likely progenitor scenario for these SNe. To constrain hydrogen in late time spectra, we present ESO/VLT and ESO/NTT optical and infrared observations of SNe Ia 1998bu and 2000cx 251-388 days after maximum. We see no hydrogen line emission in SNe 1998bu and 2000cx at these epochs, and we argue from modeling that the mass of such hydrogen-rich gas must be less than 0.03 solar masses for both supernovae. Comparing similar upper limits with recent models of Pan et al., it seems hydrogen-rich donors with a separation of less than 5 times the radius of the donor may be ruled out for the five SNe Ia 1998bu, 2000cx, 2001el, 2005am and 2005cf.
    Full-text · Article · Jul 2013 · Monthly Notices of the Royal Astronomical Society
  • A. Berkner · P. H. Hauschildt · E. Baron
    [Show abstract] [Hide abstract]
    ABSTRACT: Aims: We use our 3D radiative transfer framework to investigate how the presence of a parametrized starspot affects radiative transfer in stellar atmospheres in general, and molecular CO lines in a stellar spectrum, in particular. Methods: The equation of state is solved for a given temperature structure including a simple parametrized spot model and the 3D scattering problem for line transfer is solved via an operator splitting technique. The formal solution is based on a full characteristics solution. We have used both a LTE model and a test model with a 2 level atom, simulating a single spectral line in NLTE. Results: We present the resulting CO band spectra showing both surface resolved and an integrated total emergent flux for the star and compare the umbral, penumbral, and quiet spectrum for a solar type star, which clearly shows the presence of spots as an increased CO line-depth. Furthermore, we show that the opacity structure of the spot has a significantly different angular variation than the quiet plane parallel star and its visible shape is strongly influenced by scattering, where strong scattering disconnects the lateral intensity profile from the temperature profile of the spot. Conclusions: Even a simple, small scale parametrized model shows significant 3D effects, in both the resolved and in the surface integrated spectrum. The 3D model allows for a much more detailed treatment than simple mixing of spectra with different effective temperatures.
    No preview · Article · Feb 2013 · Astronomy and Astrophysics
  • Bin Chen · Xinyu Dai · E. Baron
    [Show abstract] [Hide abstract]
    ABSTRACT: Recent quasar microlensing observations have constrained the X-ray emission sizes of quasars to be about 10 gravitational radii, one order of magnitude smaller than the optical emission sizes. Using a new ray-tracing code for the Kerr spacetime, we find that the observed X-ray flux is strongly influenced by the gravity field of the central black hole, even for observers at moderate inclination angles. We calculate inclination-dependent flux profiles of active galactic nuclei in the optical and X-ray bands by combining the Kerr lensing and projection effects for future reference. We further study the dependence of the X-ray-to-optical flux ratio on the inclination angle caused by differential lensing distortion of the X-ray and optical emission, assuming several corona geometries. The strong lensing X-ray-to-optical magnification ratio can change by a factor of ~10 for normal quasars in some cases, and a further factor of ~10 for broad absorption line (BAL) quasars and obscured quasars. Comparing our results with the observed distributions in normal and BAL quasars, we find that the inclination angle dependence of the magnification ratios can significantly change the X-ray-to-optical flux ratio distributions. In particular, the mean value of the spectrum slope parameter αox, 0.3838log F 2 keV/F 2500 Å, can differ by ~0.1-0.2 between normal and BAL quasars, depending on corona geometries, suggesting larger intrinsic absorptions in BAL quasars.
    No preview · Article · Jan 2013 · The Astrophysical Journal
  • D. Jack · P. H. Hauschildt · E. Baron

    No preview · Article · Jan 2013 · Astronomy and Astrophysics
  • Source
    Bin Chen · Xinyu Dai · Eddie Baron · Ronald Kantowski
    [Show abstract] [Hide abstract]
    ABSTRACT: Recent quasar microlensing observations have constrained the sizes of X-ray emission regions to be within about 10 gravitational radii of the central supermassive black hole. Therefore, the X-ray emission from lensed quasars is first strongly lensed by the black hole before it is lensed by the foreground galaxy and star fields. We present a scheme that combines the initial strong lensing of a Kerr black hole with standard linearized microlensing by intervening stars. We find that X-ray microlensed light curves incorporating Kerr strong gravity can differ significantly from standard curves. The amplitude of the fluctuations in the light curves can increase or decrease by ~0.65-0.75 mag by including Kerr strong gravity. Larger inclination angles give larger amplitude fluctuations in the microlensing light curves. Consequently, current X-ray microlensing observations might have under or overestimated the sizes of the X-ray emission regions. We estimate this bias using a simple metric based on the amplitude of magnitude fluctuations. The half light radius of the X-ray emission region can be underestimated up to ~50% or overestimated up to ~20%. Underestimates are found in most situations we have investigated. The only exception is for a disk with large spin, radially flat emission profile, and observed nearly face on, where an overestimate is found. Thus, more accurate microlensing size constraints should be obtainable by including Kerr lensing. The caustic crossing time can differ by months after including Kerr strong gravity. A simultaneous monitoring of gravitational lensed quasars in both X-ray and optical bands with densely sampled X-ray light curves might reveal this feature. We conclude that it should be possible to constrain important parameters such as inclination angles and black hole spins from combined Kerr and microlensing effects.
    Full-text · Article · Nov 2012 · The Astrophysical Journal
  • Source
    Bin Chen · Xinyu Dai · Eddie Baron
    [Show abstract] [Hide abstract]
    ABSTRACT: Recent quasar microlensing observations have constrained the X-ray emission sizes of quasars to be about 10 gravitational radii, one order of magnitude smaller than the optical emission sizes. Using a new ray-tracing code for the Kerr space-time, we find that the observed X-ray flux is strongly influenced by the gravity field of the central black hole, even for observers at moderate inclination angles. We calculate inclination-dependent flux profiles of active galactic nuclei in the optical and X-ray bands by combining the Kerr lensing and projection effects for future references. We further study the dependence of the X-ray-to-optical flux ratio on the inclination angle caused by differential lensing distortion of the X-ray and optical emission, assuming several corona geometries. The strong lensing X-ray-to-optical magnification ratio can change by a factor of ~10 for normal quasars in some cases, and another factor of ~10 for broad absorption line quasars (BALs) and obscured quasars. Comparing our results with the observed distributions in normal and broad absorption line quasars, we find that the inclination angle dependence of the magnification ratios can change the X-ray-to-optical flux ratio distributions significantly. In particular, the mean value of the spectrum slope parameter $\alpha_{ox},$ $0.3838\log F_{2 keV}/F_{2500 {\AA}}$, can differ by ~0.1-0.2 between normal and broad absorption line quasars, depending on corona geometries, suggesting larger intrinsic absorptions in BALs.
    Full-text · Article · Nov 2012
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: 3-D astrophysical atmospheres will have random velocity fields. We seek to combine the methods we have developed for solving the 1-D problem with arbitrary flows to those that we have developed for solving the fully 3-D relativistic radiative transfer problem in the case of monotonic flows. The methods developed in the case of 3-D atmospheres with monotonic flows, solving the fully relativistic problem along curves defined by an affine parameter, are very flexible and can be extended to the case of arbitrary velocity fields in 3-D. Simultaneously, the techniques we developed for treating the 1-D problem with arbitrary velocity fields are easily adapted to the 3-D problem. The algorithm we present allows the solution of 3-D radiative transfer problems that include arbitrary wavelength couplings. We use a quasi-analytic formal solution of the radiative transfer equation that significantly improves the overall computation speed. We show that the approximate lambda operator developed in previous work gives good convergence, even neglecting wavelength coupling. Ng acceleration also gives good results. We present tests that are of similar resolution to what has been presented using Monte-Carlo techniques, thus our methods will be applicable to problems outside of our test setup. Additional domain decomposition parallelization strategies will be explored in future work.
    Full-text · Article · Oct 2012 · Astronomy and Astrophysics

Publication Stats

6k Citations
898.61 Total Impact Points

Institutions

  • 1970-2015
    • University of Oklahoma
      • Homer L. Dodge Department of Physics and Astronomy
      Norman, Oklahoma, United States
  • 2008-2013
    • University of California, Berkeley
      • Department of Physics
      Berkeley, California, United States
  • 2007
    • Astronomical Observatory Belgrade
      Beograd, Central Serbia, Serbia
  • 2001
    • Lawrence Berkeley National Laboratory
      • Physics Division
      Berkeley, California, United States
  • 1996
    • Arizona State University
      • School of Earth and Space Exploration
      Tempe, Arizona, United States
  • 1995
    • University of Bologna
      Bolonia, Emilia-Romagna, Italy