X-ray Spectra of High-z, Radio-quiet Quasars
This thread illustrates the use of the Chandra Source Catalog (CSC) in the context of a simple research project: creating a sample of X-ray spectra of high-redshift, radio-quiet quasars.
Last Update: 9 Sep 2013 - Rewrote CIAO section and general cleanup.
- Science question
- Creating a source list
- Crossmatching the sources against the CSC
- Retrieving CSC source data
- Analyzing CSC data with CIAO
X-ray spectroscopy of high-redshift quasars can help determine whether or not there is evolution in the structure of quasars and their environment as a function of epoch. Radio-quiet quasars dominate the quasar population, and have X-ray spectra uncontaminated by components connected to radio emission. In this thread, we illustrate the use of the CSC in obtaining Chandra X-ray spectra and detailed information on a small set of known sources: optically selected, high-redshift, radio-quiet quasars (high-z RQQs).
Creating a source list
We begin by compiling a preliminary list of known intermediate- and high-redshift quasars and determine which objects have suitable Chandra observations. The catalog can be used to filter the list, and to retrieve pre-calculated spectral data products and source properties for the selected sources.
For our preliminary source list, we choose a lower redshift cutoff of 1.5. We shall require that good optical, radio, and X-ray data exist, and therefore we apply apparent flux cutoffs in the optical and radio bands.
To create the source list, we query an astronomical database such as Vizier for all quasars/AGN which meet the following criteria: z>1.5, V<18, and f(5GHz) [=radio]<0.1 Jy. The targets are chosen from high-redshift surveys based on optical luminosity, and subsequently confirmed as radio-quiet by comparison with radio surveys.
In the this thread we will consider a subset of the quasars found in the search:
- BRI 0103+0032
- BRI 0241-0146
- BRI 0401-1711
- BRI 1033-0327
- BRI 2212-1626
- PSS 1443+2724
Crossmatching the sources against the CSC
We are now ready to query the CSC for X-ray observations of the sources in our list.
To create a Chandra source list matching our optically selected list, we query the catalog by source position via the CSCview Crossmatch feature. CSCview is a GUI which provides direct access to the contents of the catalog via user-specified queries.
To begin our query, we first clear any example entries which may be present in the interactive windows of the Query tab by selecting the "File->New->Empty Form" menu option - or by highlighting these items and selecting the "-" button next to the appropriate window. (You may choose to have the query form appear empty upon startup, or populated with an example query with the "Startup Query" option in "Edit->Preferences"; the latter is the default startup option.) We upload a table containing two columns of data, one each for the RA and Dec. positions of our sources in decimal degree units, from a TSV format file using the User Table -> Local File option of the CSCview crossmatch (VOTable format is also accepted). (Note the caveat associated with the User Table -> Received Table crossmatch option.) We change the default search radius (Radius parameter) for the crossmatch query from 3 arcminutes to 1 arcminute, since we expect our sources to be on-axis, and therefore truly point-like (far off-axis point sources can actually be extended). (Note that the ICRS RA and Dec of each source in the CSC are determined with an absolute uncertainty that does not exceed 5.0 arcseconds (1σ) for an isolated point source with at least 30 counts located within 5.0 arcminutes of the optical axis.) Finally, we set the Ra and Dec crossmatch parameters to the names of the corresponding columns in our input tables, and use the default values for the Sigma position errors and Obect ID string identifiers for the sources in our list. The optional crossmatch parameters and their default settings are described below:
- Radius - a single entered radius, or a selected column of radii from the user-input table, in arcsecs/arcminutes within which to search around each input source position for a CSC source match; default value is 3 arcminutes.
- Sigma - source position sigma error value(s) to assume for input sources, either a single entered value to apply to all sources in the list, or a selected column of errors from the user-input table. For example, a value of 1.0 means that the crossmatch search will assume that each of the source positions in the input list have associated 1-sigma source position errors. Default setting is no user-input source position errors.
- Object ID - the string to use to identify each source in the user-input list in the returned table of crossmatch search results, either default 'rowindex' or a selected column from the user-input table; default will return sources labeled as "row 1", "row 2", and so on.
For help with the CSCview crossmatch feature and general CSCview functionality, refer to the CSCview Help page.
After entering the crossmatch search parameters, we fill the Result Set window of CSCview with the source properties we would like returned (note that entering a crossmatch query automatically enters the Object ID associated with our sources into the Result Set window, along with the separation in arcseconds of each returned source match and a probability measure of the match). In addition to the requisite 'dataset_id' source property for accessing data products associated with sources returned by the search, we decide to retrieve the per-observation X-ray source position, broad band energy flux and counts, power-law model fit flux and parameters, and the True/False value for the "extent" flag. We also select the best estimates of these source properties (CSC "master source properties") in the event that multiple observations of the source are returned. Finally, we specify in the Search Criteria window an additional constraint (in addition to the crossmatch) - that the source(s) returned by the query do not suffer from pileup. The completed query form is shown in Figure 3.
For brief definitions of each source property contained in the catalog, refer to the Master Sources Table and Source Observations Table; for high-level explanations, see the Catalog Descriptions pages.
We may now retrieve the CSC counterparts to the sources in our preliminary list by sumbitting the crossmatch query.
Retrieving CSC source data
Saving search results
Downloading data products
Saving search results
Once a CSCview query is submitted, the Results tab opens with a table of search results and a list of associated data products. In Figure 4, we see the list of all Chandra source matches located within 1 arcminute of each of our input source positions, with an associated separation in arcseconds, and a probability value describing the confidence of the match. A probability of 1.0 means that the CSC source returned for the corresponding source in the input list is an exact match (down to many significant digits in the source position), and a probability of 0.0 means it is very unlikely that it is a true match. We see that multiple source matches are returned for the first and fifth sources in our input list; the best match of all CSC sources returned for a single source in an input list is the CSC source with the highest probability value associated with it.
Figure 4. CSCview Results tab, displaying a table of source properties for the sources in the preliminary list
BRI 0103+0032 -> CXO J010619.2+004823 BRI 0241-0146 -> CXO J024401.8-013403 BRI 0401-1711 -> CXO J040356.6-170322 BRI 1033-0327 -> no CXO match BRI 2212-1626 -> CXO J221527.2-161133 PSS 0059+0003 -> CXO J005922.6+000301
We notice that a search for the source BRI 1033-0327 (row 2 in our input list) does not return an X-ray object - this doesn't necessarily mean that the source was too faint to be observed by Chandra. The source might have been detected, but it could have failed to satisfy quality assurance (QA) and catalog inclusion (CI) filters in catalog processing; in particular, the CI filter imposes a source significance threshold. It is also important to keep in mind that the CSC is constructed from pointed observations; it is not an all-sky catalog, and does not include sources detected to a uniform depth. The first release of the catalog includes only point and compact sources, with observed (i.e., un-deconvolved) spatial extents <~30 arcsec; sources larger than this will not be detected with the current CSC algorithms. In addition, observations of fields containing extended sources have been excluded from the catalog, or in some cases only a part of the field has been included. More information about the processing footprint of individual observations contained in the catalog is forthcoming.
As we scroll through the columns of the results table, we note that several cells are empty for some of our sources, e.g. the power-law model fit parameters and flux (the CSC does not record spectral fit properties for sources with less than 150 counts). Instead of leaving these spaces blank, we choose to enter vales of "NULL" so that a saved table of query results will not contain blank spaces for these properties; this is done via the Edit-->Preferences option in CSCview before saving the file. Then, we can save the table of source properties to a TSV (or VOTable) format file by clicking Save in the Results tab of CSCview.
The contents of the CSC save file would appear as follows, with a brief description of each selected catalog column recorded in the header:
unix: cat csc_quasar1_results.tsv #Column separation (F9.2) Distance from source to center of cone search (.separation) ["arcsec"]  #Column name (A20) Source name in the format 'CXO Jhhmmss.s +/- ddmmss' (master_source.name) [""] [meta.id;meta.main] #Column obsid (I5) Observation identifier (obi_source.obsid) [""] [meta.id;obs.param] #Column obi (I3) Observation interval number (obi_source.obi) [""] [meta.id;obs.param] #Column ra (F9.5) Source position, ICRS right ascension (master_source.ra) ["deg"] [pos.eq.ra;meta.main] #Column ra_b (F9.5) Source position, ICRS right ascension; ACIS broad energy band (obi_source.ra_b) ["deg"] [pos.eq.ra;em.X-ray] #Column dec (F9.5) Source position, ICRS declination (master_source.dec) ["deg"] [pos.eq.dec;meta.main] #Column dec_b (F9.5) Source position, ICRS declination; ACIS broad energy band (obi_source.dec_b) ["deg"] [pos.eq.dec;em.X-ray] #Column livetime (F9.1) Effective exposure time after applying the good time intervals and the deadtime correction factor (obi_source.livetime) ["s"] [time.duration;obs.exposure] #Column significance (F7.2) Highest source flux significance across all observations (master_source.significance) [""] [stat.snr] #Column flux_significance_b (F9.2) Significance of the source determined from the ratio of the source flux to the estimated error in the local background; ACIS broad energy band (obi_source.flux_significance_b) [""] [stat.snr;em.X-ray] #Column nh_gal (G9.4) Galactic neutral Hydrogen column density, N_H(Gal) in the direction of the source determined from Dickey & Lockman (Dickey, J. M., & Lockman, F. J. 1990, ARA&A, 28, 215) (master_source.nh_gal) ["10^20/cm^2"] [phys.absorption.gal;phys.columnDensity] #Column alpha (F9.2) Photon index (alpha, defined as F_E ~ 1/E^alpha) of the best power-law model spectral fit to the source region aperture PI spectrum (master_source.alpha) [""] [spect.index;stat.fit.param;meta.modelled;em.X-ray] #Column alpha (F9.2) Photon index (alpha, defined as F_E ~ E^-alpha) of the best power-law model spectral fit to the source region aperture PI spectrum (obi_source.alpha) [""] [spect.index;stat.fit.param;meta.modelled;em.X-ray] #Column flux_powlaw (E9.3) Net integrated 0.5-10 keV energy flux of the best power-law model spectral fit to the source region aperture PI spectrum (master_source.flux_powlaw) ["erg/s*cm^2"] [phot.flux;stat.fit.param;meta.modelled;em.X-ray] #Column flux_powlaw (E9.3) Net integrated 0.5-10 keV energy flux of the best power-law model spectral fit to the source region aperture PI spectrum (obi_source.flux_powlaw) ["erg/s*cm^2"] [phot.flux;stat.fit.param;meta.modelled;em.X-ray] #Column flux_aper_b (E9.3) Aperture-corrected net energy flux inferred from the source region aperture, calculated by counting X-ray events; ACIS broad energy band (master_source.flux_aper_b) ["erg/s*cm^2"] [phot.flux;src.net;em.X-ray] #Column flux_aper_b (E9.3) Aperture-corrected net energy flux inferred from the source region aperture, calculated by counting X-ray events; ACIS broad energy band (obi_source.flux_aper_b) ["erg/s*cm^2"] [phot.flux;src.net;em.X-ray] #Column src_cnts_aper_b (G9.5) Aperture-corrected net counts inferred from the source region aperture; ACIS broad energy band (obi_source.src_cnts_aper_b) ["counts"] [phot.count;src.net;em.X-ray] #Column extent_flag (A5) Deconvolved source extent is inconsistent with a point source at the 90% confidence level (master_source.extent_flag) [""] [meta.code;phys.angSize] #Column extent_code (I3) Deconvolved source extent is inconsistent with a point source at the 90% confidence level (bit-coded value) (obi_source.extent_code) [""] [meta.code;phys.angSize] #Column hard_hs (F9.4) Spectral hardness ratio measured between ACIS energy bands 'h' and 's'; hard_hs = (flux_aper_h - flux_aper_s)/flux_aper_b (master_source.hard_hs) [""] [phot.color;em.X-ray] #Column hard_hs (F9.4) Spectral hardness ratio measured between ACIS energy bands 'h' and 's'; hard_hs = (flux_aper_h - flux_aper_s)/flux_aper_b (obi_source.hard_hs) [""] [phot.color;em.X-ray] #Column posid (I7) Internal identifier for a Source Observation (obi_source.posid) [""] [meta.id] #Column dataset_id (I10) Dataset identifier used to access archive files (obi_source.dataset_id) [""] [meta.id] separation name obsid obi ra ra_b dec dec_b livetime significance flux_significance_b nh_gal alpha alpha flux_powlaw flux_powlaw flux_aper_b flux_aper_b src_cnts_aper_b extent_flag extent_code hard_hs hard_hs posid dataset_id 0.12 CXO J010619.2+004823 2180 0 16.58018 16.58018 0.80650 0.80650 3709.2 4.78 4.78 3.160 3.005e-14 3.005e-14 24.524 FALSE 0 -0.2773 -0.2773 93928 93928 ...
Downloading data products
Level 3 data products, such as images, event lists, and spectra, may be downloaded for each source contained in the CSC from the Products tab of CSCview (see the Data Products page for the full list of data files provided by the CSC). At this point, we can download the X-ray PHA spectra, ARF and RMF response files for our sample of optically selected quasars with corresponding Chandra observations. In the Results tab of CSCview, we select the row of data in the table of source properties and the desired files from the list of data products, then click "Search". See the CSCview thread Retrieving Data Products for details on download options.
The downloaded spectrum (pha3.fits file) may be loaded into Sherpa and visualized:
unix% sherpa ----------------------------------------------------- Welcome to Sherpa: CXC's Modeling and Fitting Package ----------------------------------------------------- CIAO 4.2 Sherpa version 2 Tuesday, July 6, 2010 sherpa> load_pha("acisf02180_000N001_r0002_pha3.fits") read ARF file acisf02180_000N001_r0002_arf3.fits read RMF file acisf02180_000N001_r0002_rmf3.fits read background file acisf02180_000N001_r0002_pha3.fits sherpa> calc_data_sum() # total counts 26.0 sherpa> group_counts(1) # require minimum of 1 count per bin sherpa> prefs = get_data_plot_prefs() sherpa> prefs["yerrorbars"]=0 sherpa> plot_data()
The FITS format PHA file records the low resolution PI spectrum of the events extracted from the source region and background region, in separate FITS HDUs. A plot of the grouped source spectrum is shown in Figure 5.
Analyzing CSC data with CIAO
Fitting source spectra with Sherpa
Plotting source properties with ChIPS
Using save files and data products with CIAO
Armed with tables of CSC source data and spectral PHA files, we can begin an analysis of high-z RQQs in the X-ray regime.
Fitting CSC source spectra with Sherpa
We can use Sherpa to conduct a spectral analysis of the X-ray source regions matching our list of optically selected quasars; see the Sherpa threads page for a full list of spectral analysis options. Noting that the CSC does not contain power-law model spectral fit fluxes for sources which, like ours, have <150 counts, we can use Sherpa to manually derive meaningful spectral fits to our data.
Each Chandra Level=3 PHA file (pha3.fits) available in the Chandra Source Catalog contains both a source and background spectrum in separate FITS HDUs (CXC Data Model blocks). When a pha3.fits file is loaded into Sherpa with load_data or load_pha, the background spectrum is automatically recognized and read in as well, with the same filename as the source spectrum. The resulting source and background Sherpa data sets may be handled in the usual way in Sherpa.
Here, we fit an absorbed 1-D power law model to one of our PHA spectra downloaded from the CSC, with a Galactic neutral hydrogen column density fixed at 3e20 cm^-2 and an initial guess of 2 for the power law photon index:
unix% sherpa ----------------------------------------------------- Welcome to Sherpa: CXC's Modeling and Fitting Package ----------------------------------------------------- CIAO 4.2 Sherpa version 2 Tuesday, July 6, 2010 sherpa> load_pha("acisf02180_000N001_r0002_pha3.fits") read ARF file acisf02180_000N001_r0002_arf3.fits read RMF file acisf02180_000N001_r0002_rmf3.fits read background file acisf02180_000N001_r0002_pha3.fits sherpa> group_counts(1) # require minimum of 1 count per bin sherpa> set_source(xsphabs.abs1*powlaw1d.p1) sherpa> abs1.nh = .03 sherpa> freeze(abs1.nh) sherpa> p1.gamma = 2 sherpa> show_source() Model: 1 (xsphabs.abs1 * powlaw1d.p1) Param Type Value Min Max Units ----- ---- ----- --- --- ----- abs1.nh frozen 0.03 0 100000 10^22 atoms / cm^2 p1.gamma thawed 2 -10 10 p1.ref frozen 1 -3.40282e+38 3.40282e+38 p1.ampl thawed 1 0 3.40282e+38 sherpa> set_stat("cstat") # appropriate for low-counts data sherpa> set_method("neldermead") sherpa> fit() Solar Abundance Vector set to angr: Anders E. & Grevesse N. Geochimica et Cosmochimica Acta 53, 197 (1989) Cross Section Table set to bcmc: Balucinska-Church and McCammon, 1998 Dataset = 1 Method = neldermead Statistic = cstat Initial fit statistic = 6.50845e+06 Final fit statistic = 28.3459 at function evaluation 295 Data points = 24 Degrees of freedom = 22 Probability [Q-value] = 0.1645 Reduced statistic = 1.28845 Change in statistic = 6.50842e+06 p1.gamma 1.96741 p1.ampl 7.96099e-06 sherpa> projection() Dataset = 1 Confidence Method = projection Fitting Method = neldermead Statistic = cstat projection 1-sigma (68.2689%) bounds: Param Best-Fit Lower Bound Upper Bound ----- -------- ----------- ----------- p1.gamma 1.90571 -0.304781 0.313394 p1.ampl 7.56736e-06 -1.40987e-06 1.60581e-06
Once we are satisifed with the power-law fit statistic and the model parameter values, we can repeat this procedure for all of our sources and record pertinent fit information for further analysis. For example, we can record the power-law photon index values obtained from Sherpa fitting to determine whether it correlates with quasar redshift.
Plotting CSC source properties with ChIPS
Assuming we have an ASCII file containing a list of quasar redshift values (obtained from optical surveys) and the corresponding power-law model photon index values we obtained with Sherpa, we can use ChIPS to plot the two columns against one another:
unix% more z_vs_xray_alpha.dat #z_em alpha alpha error #---- ----- -------- 4.437 1.78 0.30 4.053 1.33 0.36 4.236 1.28 0.37 3.99 1.27 0.37 4.178 0.75 0.41 unix% chips chips> add_curve("z_vs_xray_alpha.dat[cols #1, #2, #3]") # plot z versus alpha chips> set_curve(["line.style","none"]) # remove line connecting data points chips> set_plot_xlabel ("quasar redshift") chips> set_plot_ylabel ("X-ray power-law photon index")
The resulting plot is shown in Figure 6.
ChIPS supports the same file formats as those of the CIAO Data Model, so it can be used to plot the columns of data in our CSC save file of search results. See the ChIPS threads page for a complete list of plotting options.
Using CSC save files and data products with CIAO
The file saved from CSCView can easily be manipulated by DM-specific tools in CIAO; for an introduction to the Data Model syntax used by the CIAO tools, refer to the Data Model help page. We can now use ChIPS to plot the columns of data in this file (with the syntax used in the previous section), such as the distribution of source counts, Galactic NH, fluxes, or other source properties retrieved from the CSC. For example:
unix% ciao unix% chips chips> add_curve("q_results_out.tsv[cols significance,flux_aper_b][opt kernel=text/tsv]") # plot b-band flux significance vs. counts chips> set_curve(["line.style","none"]) # remove line connecting data points chips> set_plot_xlabel ("source b-band flux significance") chips> set_plot_ylabel ("0.5-7.0 keV source counts")
To perform other basic X-ray imaging, timing, or spectral analysis with the data products and save files downloaded from the CSC, we can follow the relevant CIAO science threads - keeping in mind that many of the thread steps needed to handle problematic data can be skipped because the CSC data products are uniform and up-to-date. For more information on the usage of these files, see the page Using L3 Data Products.
For instance, we may be interested in viewing or defining the spatial boundaries of each of our sources in events images, to search for interesting emission features or even to extract a new PHA spectrum. For this purpose we could use the CIAO threads Creating Source and Background Files and Using Regions in ds9, which describe how to define source and corresponding background regions and display them on an events image in ds9. However, the CSC comes pre-packaged with a file, reg3.fits, which contains both the source and background extraction region definitions for each source; armed with this file, we can skip most of the steps contained in these threads and jump right to the punchline:
ciao% dmcopy acisf02180_000N001_r0002_reg3.fits"[SRCREG]" source.reg ciao% dmcopy acisf02180_000N001_r0002_reg3.fits"[BKGREG]" bkg.reg ciao% ds9 acisf02180_000N001_evt3.fits -region source.reg -region bkg.reg &
The last command opens an image of the full-field Level 3 events file in ds9, with the regions defined in reg3.fits displayed (see Figure 8).
Figure 8. ds9 image of L3 full-field events file (evt3.fits) with source and background regions (reg3.fits) overlaid
We can also display the regions on the source region event file, regevt3.fits:
ciao% ds9 acisf02180_000N001_regevt3.fits -region source.reg -region bkg.reg &
Figure 9. ds9 image of L3 source region events file (regevt3.fits) with source and background regions (reg3.fits) overlaid
After repeating this procedure for all the quasars in our list, we notice that none appear to be inconsistent with point sources, as suggested by the extent_flag and extent_code values of "FALSE" and "0" in our table of CSCview search results. This seems to confirm what other investigations have found, that high-z RQQs do not exhibit jets or other extended features in the X-ray regime.
This thread covers just a few examples of the many uses of CSC data products and source properties in multi-faceted, high-level scientific investigations of point sources like high-z RQQs in the X-ray regime.
|12 Feb 2009||original version|
|21 May 2009||updated for CSCview version 1.0.2|
|11 Aug 2010||updated for CSCview version 1.1|
|24 Nov 2010||updated for CSCview version 1.1.1|
|09 Sep 2013||Rewrote CIAO section and general cleanup.|