spacer link to MAST page spacer logo image spacer

NOTE: Most of the High Level Science Products are unavailable while unscheduled maintenance is being performed. They will be incrementally restored over the course of this week. We apologize for any inconvenience.

A First Catalog of Variable Stars Measured by ATLAS ("ATLAS-VAR")

Heinze et al. 2018, ApJ, submitted


Description of Data Products

Data Access

Download README


Amplitude vs. period for our PULSE variables. Color encodes φ2 - φ1, the phase offset of the first and second Fourier terms. Phase offsets > 120° (very rare in this class) have been scaled to 120° to focus on the interesting regime. The astrophysical sequences of the RRab and RRc types are clearly resolved, as are the δ Scuti variables. RRc variables have more symmetrical lightcurves, while δ Scuti and RRab stars have similarly shaped sawtooth lightcurves. The dark blue and purple points in the box at lower right are stars with phase shifts centered near 90°, indicating unusual and distinctive lightcurves with narrow, symmetrical maxima --- possibly a new class of variables.
The ATLAS-VAR data release is a catalog of variable stars discovered from ATLAS data. The first data release, presented here, consists of variable stars identified from 1.4 x 108 stars down to a limiting magnitude of r ~ 18 that obtained a minimum of 100 observations during the first two years of ATLAS operations. A total of 4.7 million variable star candidates are detected, using a Lomb-Scargle periodogram and several variability metrics described in Heinze et al. (2018). The catalog is available though the MAST CasJobs SQL interface. You can refer to the full collection of PS1COSMO data products in papers or webpages using the DOI

Description of Data Products


The catalog contains a number of tables in the "HLSP_ATLAS_VAR" Context within MAST CasJobs. The tables are:

  • detection - All ATLAS photometric measurements of all the candidate variables (991,071,517 rows).
  • object - Catalog of candidate variable stars (4,734,215 rows).
  • observation - A catalog of all ATLAS images used for DR1 (276,868 rows).

Data Access

The primary way to interact with the ATLAS-VAR catalog tables is through MAST CasJobs. For convenience, we also provide the "object" and "observation" tables as binary FITS files, but the "detection" table is only available through MAST CasJobs.

FITS File Size (GB)
Observation Table 62 MB
Object Table 7 GB

You must register for a MAST CasJobs account from there if you do not already have one. Note that MAST CasJobs accounts are separate from STScI accounts. After logging in, you can navigate to the ATLAS_VAR tables by following MyDB (in the top menu bar) → HLSP_ATLAS_VAR (in the drop-down menu at the top-left). You can click on one of the tables and then hit the Sample button to get some example rows. The column names and data types are always available in the Table Schema underneath the Notes and Sample buttons.

You can execute SQL queries by visiting the Query tab in the top menu bar. Make sure you change the Context drop-down menu to HLSP_ATLAS_VAR. We have provided a few tutorials that demonstrate how to query the ATLAS-VAR variable star catalog.

NOTE: When entering these examples into CasJobs, do NOT include the numbers or pipe characters ("|"), those are only used to reference the line numbers here.

Tutorial 1: Select 10 Targets Classified As "Pulse" With Period Between 0.3 and 0.9 Days And Return The Cyan Filter Light Curves

1| WITH ctePLUS (ATO_ID, class, fp_period, objid)

2| as

3| (select top 10 ATO_ID, class, fp_period, objid from object

4| where class='PULSE' and fp_period between 0.3 and 0.9)

5| select o.ATO_ID, o.class, o.fp_period, d.filter, d.m, d.mjd

6| from ctePLUS as o inner join detection as d on o.objid=d.objid

7| where d.filter='c'

8| order by o.ATO_ID, d.mjd

Lines 1-4 randomly selects 10 targets from the "object" table that have class="PULSE" and period between 0.3 and 0.9 days. The ATO_ID, class, fp_period, and objid columns are selected for these 10 targets and assigned to the "ctePLUS" variable.

Line 5 selects the ATO_ID, class, and fp_period values from the "ctePLUS" variable we created in Lines 1-4, and selects the filter, m, and mjd columns from the "detection" table.

Line 6 assigns the "ctePLUS" variable to "o" and the "detection" table as "d" so that they can be used as short-hand elsewhere, for example, Line 5, 7, and 8. It also performs a "join" between the "ctePLUS" variable and the "detection" table using the "objid" column from both. The "objid" column is a common key that allows SQL to combine the information in the "ctePLUS" and "detection" tables using this column as a common link between the two.

Line 7 applies a further constraint to only return the "cyan" ATLAS filter data, while Line 8 orders the output table by first the ATO_ID, then the MJD.

Tutorial 2: Do A Cone Search Centered On A Specified RA, Dec Coordinate And Return Light Curves For Both Cyan And Orange Filters

1| select o.objid, o.ra, o.dec, d.mjd, d.m,, d.filter

2| from fGetNearbyObjEq(295.85266, 19.82466, 300./3600.) as nb

3| inner join object as o on o.objid = nb.ObjID

4| inner join detection as d on o.objid = d.objid

5| order by o.objid, d.filter, d.mjd

6| into myDB.cone_search_results

Line 1 selects the columns we want to output, specifically, the object ID, RA, and Dec columns from the "observation" table and the MJD, mag, and mag uncertainty columns from the "detection" table.

Line 2 executes a cone search centered on a specified RA and Dec coordinate (in degrees). The search radius is the third argument, and must be given in degrees (in our example, the search radius is set to 300 arcseconds). The result of this cone search is given a nickname of "nb" for shorthand use later.

Line 3 performs a "join" combining the columns we want from the "observation" table with the output columns of our cone search using the object ID columns from both as a common key.

Line 4 performs another "join", this time combining the columns we want from the "detection" table with the "observation" table, again using the object ID columns from both as a common key.

Line 5 orders the output to be sorted first by object ID, then by filter, then by timestamp.

Line 6 then stores the result in your output space (MyDB) in a table called "cone_search_results". You can then examine, download, or even use the output table in other queries. You can visit the MyDB tab in the top menu and choose "MyDB" to access any output tables you've made. Make sure the Context drop-down menu in the upper left of the "MyDB" tab page is set to "MyDB" in order to see your output tables.