SNaX - A Database of Supernova X-ray Lightcurves

TL;DR

SNaX is a comprehensive, user-friendly database compiling X-ray lightcurves of young supernovae, enabling data search, visualization, comparison, and analysis, with ongoing expansion to include all available supernova data.

Contribution

This work introduces the first extensive, accessible database of supernova X-ray lightcurves with tools for analysis and community data submission, expanding beyond initial Type IIn focus.

Findings

Includes data for multiple supernova types

Provides customizable plotting and decay fitting tools

Enables direct comparison of supernova X-ray lightcurves

Abstract

We present the Supernova X-ray Database (SNaX), a compilation of the X-ray data from young supernovae (SNe). The database includes the X-ray flux and luminosity of young SNe, days to years after outburst. The original goal and intent were to present a database of Type IIn SNe. After having accomplished this we are slowly expanding it to include all SNe for which published data are available. The interface allows one to search for SNe using various criteria, plot all or selected data-points, and download both the data and the plot. The plotting facility allows for significant customization. There is also a facility for the user to submit data that can be directly incorporated into the database. We include an option to fit the decay of any given SN lightcurve with a power-law. The database includes a conversion of most datapoints to a common 0.3-8 keV band so that SN lightcurves may be directly compared with each other. A mailing list has been set up to disseminate information about the database. We outline the structure and function of the database, describe its various features and outline the plans for future expansion.

Full text

astronomical databases: miscellaneous; catalogs; supernovae: general; X-rays: general; X-rays: stars

1 Introduction

Young Supernovae (SNe) are objects that are observable over almost the entire wavelength range, from radio to gamma-rays111No young SN has been observed in gamma-rays, and only upper limits exist (H.E.S.S. Collaboration et al., 2015). Nonetheless it is expected that that they do emit at $\gamma$-ray wavelengths, but are beyond the detection limits of current telescopes (Kirk et al., 1995; Tatischeff, 2009; Dwarkadas, 2013; Marcowith et al., 2014), days to years after the SN explosion. Observations at optical wavelengths have been the norm in the last 400 years since the invention of the telescope. Observations at X-ray and radio wavelengths were generally not possible until the modern era; advances in technology and instrument building in the last 50 years, coupled with the launch of space-based observatories, have led to a significant increase in X-ray observations of supernova remnants (SNRs) and young SNe, measuring X-ray emission days, years, or even decades after explosion. The increased sensitivity of X-ray observatories like Chandra allows us to delineate a young SN from its neighbors with arc-second precision, and provides us with high resolution spectra of the same. For SNe in not so crowded fields XMM-Newton is an equally good option, while the Swift telescope has become the go-to instrument for initial detection. More recently, NuStar has provided a reasonably high resolution instrument to detect and obtain spectra of objects in the range of 6-79 keV for the first time ever, opening up new windows for observations of young SNe.

With the introduction of modern-day X-ray observatories the number of young X-ray SNe detected is steadily growing. The first SN to be detected in X-rays was SN 1980K, using the Einstein observatory (Canizares et al., 1982). This was followed by the X-ray detection of SN 1987A, initially with Ginga, and then with every other X-ray satellite that has operated after the time of explosion. Due to its proximity, SN 1987A has been extremely well-observed in X-rays, and has the most detailed and complete X-ray light curve. SN 1986J, SN 1978K and SN 1993J round out the first five (Schlegel, 1995) detected X-ray SNe. The pace has certainly picked up since then, especially with the launch of Chandra and XMM-Newton in 1999, followed by Swift. In 2003, Immler & Lewin (2003) counted 15 X-ray SNe. The number had grown to 25 by 2006 (Schlegel, 2006), and 32 by 2007 (Immler, 2007). In 2012, Dwarkadas & Gruszko (2012) collected together known lightcurves of most SNe, and listed 42 distinct X-ray detected SNe. This number has now grown to over 60. We note that as of the writing of this paper, all the known X-ray detected SNe are core-collapse SNe arising from massive stars. No X-ray detection of a Type Ia SN has as yet been reported.

As the rate of discovery and the number of data-points continues to grow it becomes more essential, and more difficult, to look up available information about detected X-ray SNe. Now is the time to create a well-thought-out database – while the amount of data is manageable, albeit growing quickly, rather than once the dataset becomes too large to easily assemble. There exist well maintained databases at other wavelengths, especially in the optical, such as WISeREP (Yaron & Gal-Yam, 2012). However, till now no central repository has existed for X-ray observations. Recently, Guillochon et al. (2016) have created a repository to host all available observations of SNe at all wavelengths, especially optical, X-ray and radio. This is a catalog that seeks to pull in existing data from other catalogs into one central database. While it is a commendable and ambitious endeavor, it does depend on the availability of the data already in other catalogs, or the willingness of the user community to contribute data. It has no special emphasis on the X-rays, and does not attempt to homogenize the X-ray data and present results in a common waveband for comparison as is done herein.

In this paper we announce the creation of a catalog specifically to hold X-ray observations of young SNe, the SuperNova X-ray Database, abbreviated as SNaX. SNaX can be found at the URL: http://kronos.uchicago.edu/snax. SNaX is a moderated, easy-to-use interactive database of X-ray observations of SNe. In §2 we outline the rationale for such a database and the science that can be carried out using SN X-ray lightcurves. In §3 we describe the database, its structure and contents, and the back end. In §4 we illustrate the graphical interface, the layout and access to the data. §5 distinguishes the database from similar endeavors and discusses our views for the future of the database.

2 Rationale for the SNaX database

2.1 What can one learn from X-ray lightcurves and Spectra of SNe?

Thermal X-ray emission, including thermal bremsstrahlung and line emission, depends on the square of the density of the emitting material. In a SN the emitting material is generally the post-shocked material, whose density is 4 times that of the pre-shock medium for a strong adiabatic shock. Thus, thermal X-ray emission is immediately a tracer of the density of the material into which the shock is expanding. The evolution of the X-ray lightcurve can be used to understand the density structure of the material, which can be related to the mass-loss from the progenitor star, and therefore to the SN progenitor. More sophisticated (and somewhat approximate) analysis can help to understand the medium structure even in the case of weak, radiative and/or reverse shocks (Chevalier & Fransson, 1994; Fransson et al., 1996; Dwarkadas & Gruszko, 2012; Chandra et al., 2009, 2012; Dwarkadas et al., 2016). If the emission is non-thermal, the mechanism is generally inverse Compton or synchrotron emission, as postulated for Type IIP (Chevalier et al., 2006) and Type Ib/c (Chevalier & Fransson, 2006) SNe. The X-ray lightcurves can also be used to decipher the density structure in these cases (Chakraborti et al., 2013, 2016).

These relationships mean that X-ray lightcurves can allow us to (1) distinguish between adiabatic and radiative shocks from SNe in certain circumstances, and (2) distinguish between expansion into a steady or non-steady wind which may arise from the SN progenitor222As of this writing, X-ray emission has been detected only from core-collapse SNe, which arise from massive stars that have strong winds that modify their circumstellar environment. Upper limits are available for Type Ia SNe (Russell & Immler, 2012; Margutti et al., 2014). In some cases the winds may create circumstellar wind-blown bubbles, further complicating the environment (Dwarkadas, 2005), and having significant implications for the X-ray emission. Lightcurves further (3) allow us to derive the slope of the density profile of the medium into which the SN is expanding, (4) reveal details about the SN kinematics, dynamics and evolution, when coupled with spectral information and other multi-wavelength data, and (5) provide valuable information regarding elemental abundances, both in the shocked circumstellar medium formed by the SN progenitor and in the SN ejecta.

2.2 Goals of the SNaX database

It is clear that there is a need to compile the X-ray fluxes of young SNe, which has led to the creation of the SNaX database. The main goals for the creation of the SNaX database are the following:

To have a moderated database which will (eventually) present all available data on X-ray SNe. 2. 2.

To provide an easily accessible and searchable user interface to rapidly find the X-ray data on any given SN. 3. 3.

To provide a template that allows users to easily submit data to the database, so that it can be stored and retrieved. 4. 4.

To provide a graphical user interface enabling the user to plot the data and download plots. 5. 5.

To provide X-ray fluxes and luminosities in a standard waveband so that all the data can be compared. One of the problems outlined in Dwarkadas & Gruszko (2012) was that it was very difficult to compare lightcurves of various SNe, or indeed even create a lightcurve for a given SN, because different authors provide fluxes and/or luminosities in different wavebands. We have found fluxes for different SNe quoted in various bands, including: 0.2-2 keV, 0.4-2 keV, 0.3-8 keV, 0.4-2.4 keV, 0.5-4 keV, 0.5-8 keV, 2-8 keV, 3-8 keV, and 2.4-10 keV. SNe discovered with Swift have their flux generally quoted in the 0.2-10 keV range. This makes it extremely difficult to make quantitative comparisons between the data from various SNe. Thus, although the first step is to present all the data, an essential component is to provide X-ray flux values in a given common band. We have chosen the 0.3-8 keV band for this purpose, which is relevant to Chandra, XMM-Newton and Swift data. We urge authors who study X-ray SNe to adopt the same philosophy and provide fluxes in the 0.3-8 keV band. 6. 6.

To intentionally maintain a moderated database. We feel that this is necessary in order to maintain data integrity, evaluate submissions, and provide fluxes in a standard waveband so that they can be easily compared.

3 Structure of the Database

The data are stored in a MySQL relational database. A web interface has been developed using a combination of php and javascript, with graphing functionality provided by the open source flot graphing library (which is in turn built upon the jquery javascript library), available at http://www.flotcharts.org/. The use of MySQL allows for rapid data retrieval, and scales smoothly with an increase in the number of X-ray SNe. Only the plotting and analysis features make extensive use of javascript, ensuring that a large amount of computing power is not required by users looking up data.

3.1 Database Tables

The core of the data is stored in three MySQL tables. At the ‘highest’ level is a table containing data pertaining to the supernova itself, such as distance, explosion date, and coordinates. Connected to the supernova table is a table containing information determining the source of the observation: instrument used, dates covered in the observation, exposure time, etc. A third table contains measurements taken from each observation: count rates, flux,errors, and the energy range over which the flux was measured. Each entry also has an associated citation, stored as a bibcode to easily generate a link to the NASA ADS database.

There are two further tables that are connected to the three main tables, and are used for storing data that needs arbitrary numbers of rows. The first is used to store the IDs associated with the observations; for example, a list of Chandra obsIDs. The second stores model information in fields like column density $N_{H}$ and temperature; if the model takes into account a variety of abundances they are also stored here.

3.2 Calculated Data

The database uses the stored data to calculate the age of the SN at the time of observation, and its luminosity. Although flux data stored for a single SN may originate from a variety of papers, SNaX only records a single value for object overview information, such as the object’s distance. This means that a given X-ray luminosity may differ from that listed in the publication in which the flux measurement is listed, should an alternate value for distance be used. We have tried not to be judgemental in choosing which distance value is used, but if different papers give the luminosity for an object using different distances, we need to make a choice to use one of the given distances, thus leading to a change in some of the luminosities. Generally, values recorded by SNaX coincide with those used by the more recent publication, with the rationale that authors are aware of previous measurements; similar concerns arise with age calculations based upon an explosion date. If a publication provides only the flux values we have adopted a distance from the SIMBAD (Wenger et al., 2000) or NED databases for use in luminosity calculations.

4 Data Layout and Access

Currently our goal is to include all available published data on X-ray SNe in the database, by which we mean data having an ADS entry. We also include conversions to the 0.3-8 keV energy range, which we calculate when it is not provided. We do not include any data that exists only in poster form. In some cases, although the data is published, quantitative information–such as a table of numbers–is not available even though a lightcurve is shown; in such cases it is of course impossible for us to include this in our database. Having already come across a few such publications, we urge authors to provide tables of numbers along with graphs showing the flux or luminosity lightcurve.

The database is funded by a NASA grant to study Type IIn SNe, and its original intent was to present all X-ray data on Type IIn SNe only. As such, incorporating all the available data on Type IIn SNe into the database has been our first priority. Having mostly accomplished this, and having created a working framework and interface, our ongoing goal is to expand the database to include data on all X-ray SNe, and furthermore to provide fluxes in the 0.3-8 keV band whenever possible. We realise this will take sustained time and effort, and are carefully working on this task. We are also hopeful that users will submit information on other SN types to the database, helping us in our task of making this an extremely useful resource. Some data on other SNe is already available, for example SN 1987A (Frank et al., 2016).

4.1 Homepage

Upon entering the database, the homepage initially offers the user a choice of selecting the SN via several options (Figure 1): specifying the name of the individual SN, searching by SN type, specifying the luminosity within a given range (upper and lower bounds can be individually specified), searching for a given host galaxy, explosion date, age, satellite or instrument used, and other criteria. This opens up a list of SNe (Figure 2). Once a SN or group of SNe is selected, a single click expands each SN dataset to show all the datapoints available for a particular SN. If multiple models are available giving different luminosities, or when different groups provide alternate models for the same dataset, we have included all the available models and flux parameters. Data are given in hypertext throughout the website, as a link to the ADS entry for the paper from which the information was taken. It is also possible to see the bibcode of the relevant citation by hovering the mouse over the relevant information.

4.2 Standardized Waveband for Comparison

A major goal of this database is to not only make data on X-ray SNe easily accessible, but also provide standardized data sets that can be easily compared. We have chosen to provide fluxes in the standardized range of 0.3-8 keV for each and every SN in the database. This band was chosen because it represents the usable overlapping range of the 3 current operational X-ray satellites Swift, Chandra and XMM-Newton, which account for most X-ray observations of young SNe taken at the present time. In order to do this expeditiously, we have written a program that takes the data stored in the database, runs it through a locally stored version of HEASARC PIMMS333https://heasarc.gsfc.nasa.gov/cgi-bin/Tools/w3pimms/w3pimms.pl (Mukai, 1993) and converts the energy band to 0.3-8 keV. This is done for those SNe where sufficient data is available to allow for the models to be identified and PIMMS conversion to take place. It is unable to take into account elemental abundances; when SNaX displays a conversion from PIMMs, it assumes solar abundances, even if the original measurement had various elements with abundances fit according to the chosen model. Our conversion is also unable to take into account multiple model components. These limitations are inherent to PIMMS.

In some cases we have downloaded and fitted the data, and recalculated the flux in the 0.3-8 keV band. These instances are marked, and our goal is to eventually do this for several datapoints, especially for those that have high-resolution Chandra and XMM-Newton data. We note that this may lead to an inconsistency in that the 0.3-8 keV flux may be computed from a different model fit than was used for the fluxes listed at other wavebands. While we endeavor to check that our model gives about the same flux as listed in the database at other wavelengths, this is not always possible, especially in cases where full original model details are not provided. We urge all submitters to provide data in the 0.3-8 keV band in addition to whatever other bands they may have computed the flux. We also include an option to plot only the 0.3-8 keV flux.

4.3 Plotting

The interface allows the user to easily plot all the data for a given SN or multiple SNe. Alternatively, specific datapoints can be selected from the available list for every individual SN. There is also an option to plot the recalculated 0.3-8 keV data to allow proper comparison between the lightcurves of various SNe. The plot itself is fully customizable (Figure 3), down to the size, axis range, axis type (linear or logarithmic), plot titles etc. The legend can be located at any of the four corners, and distributed over a given number of desired columns so as not to overlap any data. Both the plot and the data itself are available for download by the user.

4.4 Lightcurve Fitting

The decay of the X-ray lightcurve can provide information about the density structure of the ambient medium in which the SN is expanding (see §2). As first postulated by Chevalier (1982), and further elucidated in Fransson et al. (1996) and Dwarkadas & Gruszko (2012), if the self-similar solution for the expansion of the SN shocks with time (Chevalier & Fransson, 1994) can be assumed, the X-ray luminosity evolves as a power-law with time. The SNaX database incorporates an option to compute the slope of the lightcurve, assuming that the luminosity (‘L’) varies with time (‘t’) as a power-law ($L\propto t^{m}$). In order to calculate the power law index, SNaX runs a linear least squares regression on the site, using the method outlined in Numerical Recipes in C (Press et al., 1992). After taking the natural logarithm of both sides, a linear fit is used to find the slope ‘m’ and intercept ‘b’ of the expression ${\rm ln}(L)=m({\rm ln}\,t)+b$. The final fit is displayed on the screen at the top right of the plot (Figure 4). When the flux values have known uncertainties $\sigma_{i}$ this regression minimizes the value of $\sum_{i=1}^{N}(\frac{y_{i}-b-mx_{i}}{\sigma_{i}})^{2}$ (for N data points). The fitting algorithm assumes symmetric Gaussian error; when the error is asymmetric SNaX is programmed to take the lower error bound to be the value of $\sigma_{i}$. Upper limits are ignored when computing the fit.

4.5 Data Uploads and Downloads

Data can be quickly viewed on the site, and lightcurves plotted for easy comparisons. If desired the data may be easily downloaded for off-line use in a comma-separated-variable (csv) format (which can be opened by almost any spreadsheet software, such as Excel) or tab-separated-variable (tsv) format which can be opened using a text editor.

We have included a facility for the user to upload data to the database. Using a web form, it is possible to submit data pre-formatted into a csv file. An excel spreadsheet-like template provides the user with a sample of the layout of the data. The submitted file is parsed by the host, and a preview of the data as it would appear in the database is displayed as long as javascript is being utilized. If the user accepts the preview and submits, an automatic email reply is generated. The data are temporarily stored in a duplicate form of the database structure, awaiting review from a moderator. If the data have been correctly parsed by the system they are transferred to the main database for access; if possible, conversion of flux data to the 0.3-8 keV band will happen at this time.

We have also included a basic API, to provide users with the ability to integrate custom applications with the SNaX database. Any database search query that can be generated using the primary html interface can be duplicated through this API, which provides data formatted using JSON. A more detailed description of the functionality provided is available at www.kronos.uchicago.edu/snax/apiDocumentation, and the interface itself is at www.kronos.uchicago.edu/snax/urlQuery.

4.6 Requirements and Testing

SNaX has been tested on at least the following systems: Windows (7, 8.1 and 10), MacOS (10.6.8, 10.10.5), and Linux (Ubuntu 12.04, Ubuntu 16.04, CentOS 7). In all cases it has been found to work properly, and perform as desired, using Firefox and Google Chrome browsers. Safari on newer systems works fine, although on an older Mac using 10.6.8, Safari 5.1 was found to have problems. Performance was as desired with the new Edge browser on Windows, and it also worked with Internet Explorer 10 on Windows 7. In general the latest versions of Firefox, Google Chrome and Windows Edge browsers should work on all systems, and meet performance expectations.

The database is accessible on mobile devices, although it is not necessarily responsive. The graph size will not adapt to the size of the browser window, and left to right scrolling may be necessary to read listed data. It has been tested on devices including an ASUS T100 tablet running Windows (with Firefox), an IPad Mini (Chrome, Safari browsers), an Amazon Kindle (Silk Browser), and an Android phone running Android 5.0.2 (Chrome, Opera Browsers). In all cases the performance was satisfactory. On devices that are unable to utilize javascript it is still possible to view and export data, although plotting and analysis options may be unavailable.

Our first implementation of the database was opened up to a few users in different countries in March 2016, who accessed a beta version and provided comments. The testers were generally satisfied with the user interface and contents, but had comments on available or additional features, search criteria, or modifications that could be made. These comments were used to make several improvements to the database, including host galaxy search and lightcurve fitting. A help section providing further usage details has been added, and will be continually updated. We think the database is now working robustly and can be opened up to astronomers worldwide. The developers welcome feedback on the structure, interface and implementation aspects of the database, that would allow us to continually improve its usefulness.

5 Discussion and Future Prospects

We have presented in this paper a new database, SNaX, that lists the X-ray data from young SNe. The database can be searched in various ways, allows for plotting of selected data or the complete dataset, and includes a functionality to upload data. All data can be downloaded if desired. Over the next few years we anticipate adding almost all available data on X-ray SNe to the database. We are hopeful that users will contribute a substantial amount of data as well to make this the go-to site for astronomers needing X-ray data on supernovae.

This database differs from other available databases, such as the Open SN Catalog, in some important ways:

• •

It allows for comparison of the lightcurves. One can plot the lightcurves of all SNe, or any specific ones as needed, on a single plot, allowing for easy comparison. To our knowledge this is not possible with other available catalogs at present.

• •

We aim to (eventually) provide data on all SNe in a single X-ray band (0.3-8 keV) so as to enable proper comparison. In the past this was impossible, as can be seen from the published datapoints, which are given in a multitude of wavebands.

• •

The moderators (currently this paper’s authors, although that could change in future) are actively looking through the published astrophysics literature for listed X-ray fluxes of SNe, and adding them to the database.

• •

Where available, we also store and provide model information, which is very useful for reproducing the analysis, or recalculating the flux in various bands.

• •

Submissions from other astronomers are welcome and appreciated. A template is provided to ensure that data are submitted in the correct format. The download is inspected by the moderators before being added to the database.

• •

The moderators are continually downloading and fitting data, and providing fluxes for those SNe whose data is available but not published.

• •

SNaX allows the user to fit the lightcurve of one or more SNe with the press of a single button, plot the fit and list the power-law index.

• •

It is moderated to ensure catalog integrity, prevent duplicates and ensure valid data uploads.

We have created a mailing list to post announcements and enable discussion of the database. Information and subscription details can be found at: https://lists.uchicago.edu/web/info/snax-users

One of our future goals is to include the reduced X-ray spectra of young SNe within the database. For this to happen, astronomers working in this area need to provide their spectral files in FITS format along with the data in published papers. There is not currently any provision to do so, but we hope that this changes over time, and that astronomers will start to provide the reduced spectra. In the meantime the database moderators are tasked with developing an interface to adequately deal with spectra in the FITS format. This is one of the major aspects to be explored in future.

We believe that the SNaX database can become a go-to resource for X-ray data on young supernovae. We hope to include all SNe as the number of SNe detected in X-rays increases, and cover all SN types. We encourage astronomers to use it and provide constructive feedback to the developers444The developers and moderators can be directly contacted via the email address listed in the database, snax@kronos.uchicago.edu so that we may continue to improve the infrastructure, contents and delivery. We do request that all users please cite this paper, which enables a counting of aggregate users and thus establishes the usefulness of the database, in addition to citing all sources of the data that they download.

We acknowledge comments and suggestions from the anonymous referee that helped to improve this paper in many ways. This work is supported by a NASA Astrophysics Data Analayis program grant

NNX14AR63G awarded to PI V. Dwarkadas at the University of

Chicago. We thanks Dr F. Bauer, Dr. F. Sutaria and Dr. R. Chevalier for comments on an earlier version of the database, and Dr. K. Frank for providing a table of SN 1987A data. This research has made use of NASA’s Astrophysics Data System; the SIMBAD database operated at CDS, Strasbourg, France; and the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration.