arXiv subjects
Publications and source records attributed to Daniel Dewey.
Success in the quest for artificial intelligence has the potential to bring unprecedented benefits to humanity, and it is therefore worthwhile to investigate how to maximize these benefits while avoiding potential pitfalls. This article gives numerous examples (which should by no means be construed as an exhaustive list) of such worthwhile research aimed at ensuring that AI remains robust and beneficial.
We report on the results from the analysis of our 114 ks Chandra HETGS observation of the Galactic core-collapse supernova remnant G292.0+1.8. To probe the 3D structure of the clumpy X-ray emitting ejecta material in this remnant, we measured Doppler shifts in emission lines from metal-rich ejecta knots projected at different radial distances from the expansion center. We estimate radial velocities of ejecta knots in the range of -2300 <~ v_r <~ 1400 km s^-1. The distribution of ejecta knots in velocity vs. projected-radius space suggests an expanding ejecta shell with a projected angular thickness of ~90" (corresponding to ~3 pc at d = 6 kpc). Based on this geometrical distribution of the ejecta knots, we estimate the location of the reverse shock approximately at the distance of ~4 pc from the center of the supernova remnant, putting it in close proximity to the outer boundary of the radio pulsar wind nebula. Based on our observed remnant dynamics and the standard explosion energy of 10^51 erg, we estimate the total ejecta mass to be <~ 8 M_sun, and we propose an upper limit of <~ 35 M_sun on the progenitor's mass.
We present the analysis of 21 bright X-ray knots in the Cassiopeia A supernova remnant from observations spanning 10 yr. We performed a comprehensive set of measurements to reveal the kinematic and thermal state of the plasma in each knot, using a combined analysis of two high energy resolution High Energy Transmission Grating (HETG) and four medium energy resolution Advanced CCD Imaging Spectrometer (ACIS) sets of spectra. The ACIS electron temperature estimates agree with the HETG-derived values for approximately half of the knots studied, yielding one of the first comparisons between high resolution temperature estimates and ACIS-derived temperatures. We did not observe the expected spectral evolutionpredicted from the ionization age and density estimates for each knotin all but three of the knots studied. The incompatibility of these measurements with our assumptions has led us to propose a dissociated ejecta model, with the metals unmixed inside the knots, which could place strong constraints on supernova mixing models.
Handed the baton from ROSAT, early observations of SN 1987A with the Chandra HETG and the XMM-Newton RGS showed broad lines with a FWHM of 10^4 km/s: the SN blast wave was continuing to shock the H II region around SN 1987A. Since then, its picturesque equatorial ring (ER) has been shocked, giving rise to a growing, dominant narrow-lined component. Even so, current HETG and RGS observations show that a broad component is still present and contributes 20% of the 0.5--2 keV flux. SN 1987A's X-ray behavior can be modeled with a minimum of free parameters as the sum of two simple 1D hydrodynamic simulations: i) an on-going interaction with H II region material producing the broad emission lines and most of the 3--10 keV flux, and ii) an interaction with the dense, clumpy ER material that dominates the 0.5--2 keV flux. Toward the future, we predict a continued growth of the broad component but a drop in the 0.5--2 keV flux, once no new dense ER material is being shocked. When? Time, and new data, will tell.
We have been monitoring the supernova remnant (SNR) 1987A with {\it Chandra} observations since 1999. Here we report on the latest change in the soft X-ray light curve of SNR 1987A. For the last $\sim$1.5 yr (since day $\sim$8000), the soft X-ray flux has significantly flattened, staying (within uncertainties) at $f_{\rm X}$ $\sim$ 5.7 $\times$ 10$^{-12}$ erg cm$^{-2}$ s$^{-1}$ (corresponding to $L_{\rm X}$ $\sim$ 3.6 $\times$ 10$^{36}$ erg s$^{-1}$) in the 0.5--2 keV band. This remarkable change in the recent soft X-ray light curve suggests that the forward shock is now interacting with a decreasing density structure, after interacting with an increasing density gradient over $\sim$10 yr prior to day $\sim$8000. Possibilities may include the case that the shock is now propagating beyond a density peak of the inner ring. We briefly discuss some possible implications on the nature of the progenitor and the future prospects of our {\it Chandra} monitoring observations.
SN 1996cr, located in the Circinus Galaxy (3.7 Mpc, z ~ 0.001) was non-detected in X-rays at ~ 1000 days yet brightened to ~ 4 x 10^{39} erg/s (0.5-8 keV) after 10 years (Bauer et al. 2008). A 1-D hydrodynamic model of the ejecta-CSM interaction produces good agreement with the measured X-ray light curves and spectra at multiple epochs. We conclude that the progenitor of SN 1996cr could have been a massive star, M > 30 M_solar, which went from an RSG to a brief W-R phase before exploding within its ~ 0.04 pc wind-blown shell (Dwarkadas et al. 2010). Further analysis of the deep Chandra HETG observations allows line-shape fitting of a handful of bright Si and Fe lines in the spectrum. The line shapes are well fit by axisymmetric emission models with an axis orientation ~ 55 degrees to our line-of-sight. In the deep 2009 epoch the higher ionization Fe XXVI emission is constrained to high lattitudes: the Occam-est way to get the Fe H-like emission coming from high latitude/polar regions is to have more CSM at/around the poles than at mid and lower lattitudes, along with a symmetric ejecta explosion/distribution. Similar CSM/ejecta characterization may be possible for other SNe and, with higher-throughput X-ray observations, for gamma-ray burst remnants as well.
A supernova (SN) explosion drives stellar debris into the circumstellar material (CSM) filling a region on a scale of parsecs with X-ray emitting plasma. The velocities involved in supernova remnants (SNRs), thousands of km/s, can be directly measured with medium and high-resolution X-ray spectrometers and add an important dimension to our understanding of the last stages of the progenitor, the explosion mechanism, and the physics of strong shocks. After touching on the ingredients of SNR kinematics, I present a summary of the still-growing measurement results from SNR X-ray observations. Given the advances in 2D/3D hydrodynamics, data analysis techniques, and especially X-ray instrumentation, it is clear that our view of SNRs will continue to deepen in the decades ahead.
SN1996cr is one of the five closest SNe to explode in the past 30 years. Due to its fortuitous location in the Circinus Galaxy at ~ 3.7 Mpc, there is a wealth of recently acquired and serendipitous archival data available to piece together its evolution over the past decade, including a recent 485 ks Chandra HETG spectrum. In order to interpret this data, we have explored hydrodynamic simulations, followed by computations of simulated spectra and light curves under non-equilibrium ionization conditions, and directly compared them to the observations. Our simulated spectra manage to fit both the X-ray continuum and lines at 4 epochs satisfactorily, while our computed light curves are in good agreement with additional flux-monitoring data sets. These calculations allow us to infer the nature and structure of the circumstellar medium, the evolution of the SN shock wave, and the abundances of the ejecta and surrounding medium. The data imply that SN 1996cr exploded in a low-density medium before interacting with a dense shell of material about 0.03pc away from the progenitor star. We speculate that the shell could be due to the interaction of a blue supergiant or Wolf-Rayet wind with a previously existing red supergiant (RSG) wind. The shock wave has now exited the shell and is expanding in the medium exterior to it, possibly the undisturbed continuation of the dense RSG wind. The narrow lines that earned SN 1996cr its IIn designation possibly arise from dense, shocked clumps in the CSM. Although the possibility for an LBV progenitor for this Type IIn SN cannot be completely excluded, it is inconsistent with much of the data. These calculations allow us to probe the stellar mass loss in the very last phases ($ < 10^4$ years) of a massive star's life ($> 10^6$ years), and provide another means to deducing the progenitor of the SN.
Astronomical data generally consists of 2 or more high-resolution axes, e.g., X,Y position on the sky or wavelength and position-along-one-axis (long-slit spectrometer). Analyzing these multi-dimension observations requires combining 3D source models (including velocity effects), instrument models, and multi-dimensional data comparison and fitting. A prototype of such a "Beyond-XSPEC" (Noble & Nowak, 2008) system is presented here using Chandra imag- ing and dispersed HETG grating data. Techniques used include: Monte Carlo event generation, chi-squared comparison, conjugate gradient fitting adapted to the Monte Carlo characteristics, and informative visualizations at each step. These simple baby steps of progress only scratch the surface of the computational potential that is available these days for astronomical analysis.
We present an extended analysis of the deep Chandra LETG and HETG observations of the supernova remnant 1987A (SNR 1987A) carried out in 2007. The global fits to the grating spectra show that the temperature of the X-ray emitting plasma in the slower shocks in this system has remained stable for the last three years, while that in the faster shocks has decreased. This temperature evolution is confirmed by the first light curves of strong X-ray emission lines and their ratios. On the other hand, bulk gas velocities inferred from the X-ray line profiles are too low to account for the post-shock plasma temperatures inferred from spectral fits. This suggests that the X-ray emission comes from gas that has been shocked twice, first by the blast wave and again by shocks reflected from the inner ring of SNR 1987A. A new model that takes these considerations into account gives support to this physical picture.
The flight calibration of the spectral response of CCD instruments below 1.5 keV is difficult in general because of the lack of strong lines in the on-board calibration sources typically available. We have been using 1E 0102.2-7219, the brightest supernova remnant in the Small Magellanic Cloud, to evaluate the response models of the ACIS CCDs on the Chandra X-ray Observatory (CXO), the EPIC CCDs on the XMM-Newton Observatory, the XIS CCDs on the Suzaku Observatory, and the XRT CCD on the Swift Observatory. E0102 has strong lines of O, Ne, and Mg below 1.5 keV and little or no Fe emission to complicate the spectrum. The spectrum of E0102 has been well characterized using high-resolution grating instruments, namely the XMM-Newton RGS and the CXO HETG, through which a consistent spectral model has been developed that can then be used to fit the lower-resolution CCD spectra. We have also used the measured intensities of the lines to investigate the consistency of the effective area models for the various instruments around the bright O (~570 eV and 654 eV) and Ne (~910 eV and 1022 eV) lines. We find that the measured fluxes of the O VII triplet, the O VIII Ly-alpha line, the Ne IX triplet, and the Ne X Ly-alpha line generally agree to within +/-10 % for all instruments, with 28 of our 32 fitted normalizations within +/-10% of the RGS-determined value. The maximum discrepancies, computed as the percentage difference between the lowest and highest normalization for any instrument pair, are 23% for the O VII triplet, 24% for the O VIII Ly-alpha line, 13% for the Ne IX triplet, and 19% for the Ne X Ly-alpha line. If only the CXO and XMM are compared, the maximum discrepancies are 22% for the O VII triplet, 16% for the O VIII Ly-alpha line, 4% for the Ne IX triplet, and 12% for the Ne X Ly-alpha line.
Using the Chandra/High Energy Transmission Grating Spectrometer (hereafter HETGS) we have detected Doppler motion of Capella's X-ray emission lines in the 6 -- 25AA wave-band. The observed motion follows the expected orbital motion of Capella's primary. This finding implies that the primary G8 III star, not the secondary G1 III star in the Hertzsprung gap, has been the dominant source of hot 10^{6.8} -- 10^{7}K plasma at least in the last six years. In addition, the results demonstrate the long-term stability of the HETGS and demonstrate small uncertainties of 25 and 33 km/s in the velocity determination with the HEG and MEG, respectively.
Details of the design, fabrication, ground and flight calibration of the High Energy Transmission Grating, HETG, on the Chandra X-ray Observatory are presented after five years of flight experience. Specifics include the theory of phased transmission gratings as applied to the HETG, the Rowland design of the spectrometer, details of the grating fabrication techniques, and the results of ground testing and calibration of the HETG. For nearly six years the HETG has operated essentially as designed, although it has presented some subtle flight calibration effects.
We introduce a million-second observation of the supernova remnant Cassiopeia A with the Chandra X-ray Observatory. The bipolar structure of the Si-rich ejecta (NE jet and SW counterpart) is clearly evident in the new images, and their chemical similarity is confirmed by their spectra. These are most likely due to jets of ejecta as opposed to cavities in the circumstellar medium, since we can reject simple models for the latter. The properties of these jets and the Fe-rich ejecta will provide clues to the explosion of Cas A.
We present results from in-flight calibration of the High Energy Transmission Grating Spectrometer (HETGS) on the Chandra X-ray Observatory. Basic grating assembly parameters such as orientation and average grating period were measured using emission line sources. These sources were also used to determine the locations of individual CCDs within the flight detector. The line response function (LRF) was modeled in detail using an instrument simulator based on pre-flight measurements of the grating alignments and periods. These LRF predictions agree very well with in-flight observations of sources with narrow emission lines. Using bright continuum sources, we test the consistency of the detector quantum efficiencies by comparing positive orders to negative orders.