Swift banner Italian site U.S. site U.K. site

Swift/XRT light curve Repository -- Online Documentation

Contents

Overview

This web site provides a repository of Swift/XRT light curves and images for every GRB this instrument has detected. To view a light curve you can enter the BAT trigger number or the GRB name in the search box and click "Submit". Alternatively, move your mouse over the year and click on the month in the drop-down menu that appears (requires Javascript). Thumbnail images of all GRB light curves available for that month will appear, click on one of these to view the full light curve. In Javascript-enabled browsers there is also a box to the right of the screen showing thumnail images of the 5 most-recently observed objects. These serve as links to the full results page for those bursts.

Once you have selected an object, you will be taken to the products page containing the light curve for that burst.

Bursts which the XRT did not observe or did not detect do not have thumbnail images, even if they are Swift-detected bursts.

Back to contents | Back to repository

Usage

The Swift/XRT team are happy for users to make use of these data products in their research and publications, provided that this is acknowledged. We request two forms of acknowledgement. Where the data are first introduced, please cite Evans et al.(2007, A&A, 469, 379) and Evans et al. (2009, MNRAS, 397, 1177) which describe how the data were produced. In the acknowledgements section, please use the following text: This work made use of data supplied by the UK Swift Science Data Centre at the University of Leicester.

Back to contents | Back to repository

Products Pages

When you view a light curve, the results will be presented to you on a web page: the "Products Page." Details of how these products were created are given in Evans et al., but please also see the changelog for subsequent alterations.

The products page begins with the name of the burst, some explanatory text and then possibly a warning. If there are observations of the burst on the quick-look site, Swift may still be observing the object. This means that the light curve will not be "complete"; new observations and data may be forthcoming. Thus, if quick-look data do exist, a warning informs you of this.

The products follow, presented first in graphical format. There will be (up to) five images, the first 3 of which link to the ASCII data from which they were created. These images are:

  1. Flux light curve (where available) -- the count-rate light curve (described next) converted to observed flux using a constant conversion factor derived by manual spectral analysis, taken from the GCN circulars or reports.
  2. Count-rate light curve -- showing the source count rate, after the background has been subtracted and corrections for pileup, bad columns, and the size of the source-data extraction region have been applied.
  3. "Detailed light curve" -- as above, but with the background level shown on the same plot (note that at low count rates, the net source count rate can lie below the background rate but still be a 3 sigma detection). The fractional exposure is also shown, in the bottom panel.
  4. Hardness ratio -- similar to the basic light curve, but in two bands: 0.3-1.5 keV and 1.5-10 keV. The bottom panel is the ratio: hard/soft.
  5. 3-colour deep image (where available) -- an image of the total exposure so far, split into three energy bands to provide `colours' and smoothed for aesthetic reasons. See note on colour images

Finally, there are a series of links to the data products in formats which cannot be embedded in a web page. i.e. postscript versions of the light curves, the ASCII data from which the images were created, and the source and background event lists created for the light curve. The format of the ASCII data files are as follows:

The "Basic light curve" ASCII file is a single file containing just the time, count rate and error columns. This is ready for use in QDP, thus begins with READ SERR/TERR lines (to tell qdp which columns contain errors) There are 5 columns in this file, which are:

  1. Time
  2. Time -- positive error (1-sigma)
  3. Time -- negative error (1-sigma)
  4. Source count rate
  5. Error in count rate

There are up to four datasets in the file, separated by a line of NO NO NO NO NO, (again, for compatibility with QDP). The datasets appear in the following order: Windowed Timing (WT) data; WT upper limit; Photon Counting (PC) data; PC upper limit. Depending on the object, some of these may be missing, however each dataset begins with a comment (denoted by a leading !) identifying it.

The "Detailed light curve" ASCII file is analogous to the basic one. The first three columns are again the time and its errors, however the final two columns depend on the dataset. As before, the datasets are identified by comment lines, but this time they appear in the following order: WT data, WT upper limit, WT background level, WT fractional exposure; followed by the same for PC-mode.

The "hardness ratio" ASCII file contains the same basic 5 columns as in the basic light curve: time, positive error, negative error, source rate, source error. There are six datasets in the file, again, with NO NO NO NO NO delimiters and leading comment lines. The datasets are: WT hard band, WT soft band, WT hardness ratio; then the same for PC-mode.

The "Detailed PC/WT mode data" are the raw output of the light curve generation, from which all the above files were parsed. Using something like awk or Perl, one can easily extract the data to create new plots. There is one file per mode, and one dataset per file. Where there is an upper limit, the file ends with a comment line announcing this fact. The columns are as follows:

  1. Time
  2. Time -- positive error
  3. Time -- negative error
  4. Net source count rate
  5. Error in the above
  6. Fractional exposure
  7. Background rate
  8. Error in background
  9. Total correction factor
  10. Exposure time in the bin (i.e. bin width * fractional exposure)
  11. Detection significance in this bin

The "total correction factor" is the ratio of the corrected count rate to observed count rate in this bin. The corrected count rate has been amended to account for: bad pixels/columns, pile-up and source counts landing outside the extraction region.
Sigma is the detection significance of the source in this bin (=net counts/error in bg counts). For an upper limit, the value written is the detection significance of the data which were replaced with the limit. The upper limit is always at the three-sigma level, and is calculated using the Bayesian method (see Kraft, Burrows & Nousek, 1991, ApJ, 374, 344).

The event lists, source and background lists for each mode, contain all source or background events used in the light curve creation. As well as the standard columns (TIME, position information, energy etc), the following information which may be of interest has been added:

  1. SRCRAD - Source extraction region radius (pixels).
  2. PUPRAD - Radius excluded due to pileup (arcseconds) - source event lists only.
  3. SRCAREA - Area of the source extraction region (square pixels) - background event lists only.
  4. BGAREA - Area of the background extraction region (square pixels) - background event lists only.

A note on the colour images. To create the colour image, the data are split into three energy bands: 0.3-1.2 keV, 1.2-1.8 keV and 1.8-10 keV. These are treated as the red, green and blue channels respectively and combined to give the colour image, which is then smoothed. The energy bands used were chosen based on the spectra of the GRBs observed by Swift to date, so that a typical GRB will have equal numbers of counts in the three channels. Thus, as one may intuitively expect, a comparatively soft burst will appear redder, and a hard burst bluer, in these images.

Back to contents | Back to repository

How to rebin the data

During lightcurve generation, there are 4 criteria which must be met before a bin is considered complete. These are:

  1. There must be at least X counts in the source region
  2. The bin must be at least Y seconds long
  3. The bin must use all events from a given CCD frame
  4. The source count rate must be at least 3 sigma above the background

Criterion 3 is a little obscure. Basically, since the data are read out at discrete times, one can get many events with the same timestamp. Criterion 3 ensures that these events are all put in the same bin, to prevent overlapping error bars.

In the standard products, the data are binned dynamically; that is, the binning criteria vary with count rate.
X, the minimum number of counts in a bin, is valid at a count rate of 1 count per second (cps); when the count rate changes by some factor (the rate factor), X changes by some other factor (the bin factor). The default values for these factors are 10 and 1.5 respectively, thus an order-of-magnitude change in count rate produces a factor of 1.5 change in bin size. This is done discretely rather than continuously, so if X=20 then where the count rate lies in the range 1-9.99 cps a bin must contain at least 20 counts. However, when the count rate is 10-99.99 cps, there must be 30 counts for a bin to be full. At lower rates of 0.1-0.99 cps, only 13 (=20/1.5) counts are needed to complete a bin, although this is rounded to 15: no matter how low the count rate goes, a bin will always need at least 15 counts, to ensure that using Gaussian methods to propogate errors statistics is valid.

While the above approach gives useful, usable results for almost all GRBs, it may not give the best possible results for a given burst. We have thus provided an interface to allow the user to rebin the light curves for themselves.

From the products page, follow the "Rebin this lightcurve" link. The form with which you are presented contains the fields detailed below. If you have javascript enabled, clicking on the field name will open a help box with more information about that field.

Counts per bin:
The minimum number of counts needed for a bin to be complete (criterion 1 above). Must be specified for WT and PC mode.
Minimum bin length:
The minimum duration required for a bin to be complete (criterion 2 above). Must be specified for WT and PC mode.
Dynamic binning:
Whether to use dynamic bining
Rate/bin factor:
These two boxes control the dynamic binning (if selected), full details are above.
Minimum counts per bin:
Dynamic binning cannot reduce the minimum counts per bin indefinitely. This defines the "hard" minimum, i.e. regardless of dynamic binning there must always be at least this many counts. We recommend that this is at least 15, to ensure that the Gaussian error propogation is valid.
Minimum sigma:
The minimum significance a bin must have to be considered complete (criterion 4 above).
Energy and grade selection:
The default is to use the same energy and grade ranges which were used for the light curve you are rebinning (the standard light curves all use 0.3-10 keV for the main light curve, 0.3-1.5 keV and 1.5-10 keV as the soft and hard bands respetively, and grades 0-12). You can select "User-specified" here, allowing you to set these fields yourself. If you do not have javascript, this option will not work. Instead, you should follow the link at the top of the page for advanced rebinning.

The following fields are only available if you have chosen "User specified" for the energy and grade selection.

Energy range for main curve:
Only events with energies in this range will be included in the light curves.
Grade range for main curve:
Only events with grades in this range will be included in the light curves.
Soft energy band:
The range of energies to use in the soft band for the hardness ratio. Note, these energies must lie within the range for the main curve.
Hard energy band:
The range of energies to use in the hard band for the hardness ratio. Note, these energies must lie within the range for the main curve.

Once you are happy, click "Rebin data". The rebinning process may take a while — the duration scales roughly with number of events — so for particularly bright or long lasting bursts this can take over five minutes. You will be redirected to a "holder page" which will auto-reload every 30 seconds until the process is finished at which point it will be replaced with the products page. Note: your rebinned lightcurve will be stored at a temporary URL, and deleted after 24 hours. There is no link to it, so it is a good idea to bookmark the holder page when the rebin process starts, so that you can find the page again.

Back to contents.

Updating the data

As bursts are observed, these light curves will automatically be updated when the new data are received from the spacecraft, roughly once every two hours.

Back to contents | Back to repository

New Bursts

The light curves of Swift-BAT detected GRBs will automatically be created when data are available, and placed on this website. Under certain circumstances (e.g. a very faint burst, or several candidate afterglows) there may be a slight delay in producing the curves.

Non-BAT detected bursts are not built automatically. Thus, the light curve will not be created until a member of the Swift team enters the correct details to the software. Once this is done, the curve will be made available, and updates will proceed automatically.

Back to contents | Back to repository

Changelog

The light-curve generation software is occasionally modified. Any changes are listed here.

Back to contents | Back to repository