Log-in to SkyMapper

DR4 | DR2 | DR1.1 | EDR

Data Release DR4 | 5 Feb 2024

DOI: 10.25914/5M47-S621

Browse table metadata | Documentation | Known issues with this release

💾 Download the full DR4 catalogue

Accessibility: World-wide accessible

Release documentation

DR4 Documentation

Please read and cite the DR4 paper (submitted to PASA and available on arXiv).

Introduction^ Back to top

This Fourth Data Release further enhances SkyMapper's deep coverage of the Southern sky and improves the photometric homogeneity for the u and v filters. DR4 more than doubles the image dataset compared to DR3. The data processing includes several enhancements over previous DRs. Included in DR4 are images taken between March 2014 and September 2021, with a number of quality cuts applied. Measurements are included from over 15 billion detections covering more than half the sky (field centres up to Declinations of +28°). They correspond to over 700 million unique astrophysical objects from magnitude 9 to 22. (All magnitudes reported on this site are AB mags.) The object identifiers from DR2/DR3 have been retained where appropriate, to facilitate quick checks for revised photometry. 

Please address all feedback, suggestions, and bug reports in the first instance to skymapper@anu.edu.au.


What type of data is provided?^ Back to top

DR4 contains reduced images with overscan, bias, flatfield correction and a World Coordinate System (TPV projection) applied, as well as photometric catalogues (both raw detections and object-merged) for objects in each image. Full access to the DR4 database tables is provided through this website and the Virtual Observatory Table Access Protocol (TAP).

Five SkyMapper-specific database tables have been exposed in this data release:

The dr4.master table has been pre-matched to several other large catalogues:

Matching to the external tables was attempted with a 15" search radius around each SkyMapper source and the ID and distance of the nearest source from the those catalogues are noted in sets of columns in dr4.master. As in DR2 and DR3, only the 2MASS PSC was queried (instead of both the PSC and XSC), while for Gaia DR3, the two nearest matches from the catalogue are listed. The distance to the 2nd-nearest match may indicate the presence of a close neighbour affecting the photometry of the primary source (also see the self_id1, self_id2, and self_id3 columns in dr4.master for the three closest DR4 sources to each DR4 object). The ext schema also contains additional photometric catalogues that have not been cross-matched with DR4 (see table metadata here).

We also provide local copies of several spectroscopic and other catalogues, which we have pre-matched to the SkyMapper master table with a 15" search radius, noting the nearest SkyMapper source in the dr4_id and dr4_dist columns of those tables:

To aid in constructing your ADQL queries, you can browse the column metadata of all these tables here.


Sky Coverage and Data Quality^ Back to top

DR4 covers a total area of over 26,000 deg2 (see coverage maps per filter below).

The 417,223 images in DR4 are comprised of 47% Shallow Survey (SS) and 29% Main Survey (MS) frames, and, for the first time, incorporates repeat visits to eight CALSPEC standard star fields (11%) as well as images from the SkyMapper Transient Survey (SMT; 8%; Scalzo et al. 2017) and non-survey images collected for smaller science programs (5%).

The maximum photometric zeropoint (the magnitude equivalent to one detected count in the image) is shown in each of the 6 sky maps below (in which darker colours are deeper images). The typical 5-sigma detection limits are 7-7.5 magnitudes brighter than the indicated zeropoint.

To indicate the imaging depth, the logarithmic number counts of sources with < 5% errors are shown in the figure below, with the bimodal distributions principally reflecting the Shallow Survey and Main Survey components.

Median seeing in DR4 is (3.3", 3.1", 2.9", 2.6", 2.6", 2.5") for (u, v, g, r, i, z), and is typically slightly better in Main Survey images than in the Shallow Survey. We have included images with a PSF FWHM of up to 5" and an elongation of up to 1.4.

All source detections in the photometry table include the Source Extractor flags, complemented by additional flags from our data processing in a higher bit range:

From Source Extractor:

Flag Bit Meaning
1 The object has neighbors, bright and close enough to significantly bias the photometry, or bad pixels (more than 10% of the integrated area affected).
2 The object was originally blended with another one.
4 At least one pixel of the object is saturated (or very close to).
8 The object is truncated (too close to an image boundary).
16 Object's aperture data are incomplete or corrupted.
32 Object's isophotal data are incomplete or corrupted.
64 A memory overflow occurred during deblending.
128 A memory overflow occurred during extraction.

Our post-processing adds the following flags, as needed:

Only detections with flags<4 and nimaflags<5 are used for distilling photometry into the master table, and hence the averaged magnitudes exclude all measurements with bad flags (flags>=4 or nimaflags>=5). When all detections of an object have bad flags in a passband, then the magnitude errors are nulled to indicate that the magnitudes need to be taken with a grain of salt.

However, all detections with bad flags or those close to bright stars can still be found in the photometry table.

Data Access^ Back to top

Main Page: How To Access

DR4 images and catalogues can be accessed via the tools on this website (see How To Access), which are being developed as part of the Australian All-Sky Virtual Observatory. The TAP (Table Access Protocol), SIAP (Simple Image Access Protocol), and Cone Search services can also be accessed through Virtual Observatory-aware software tools like TOPCAT and Aladin

For TAP queries, some definitions and constraints worth considering when making selections from the dr4.master catalogue include:

Caveats and Known Issues^ Back to top

For an up-to-date list of known issues and their prospects for resolution, please see the Known Issues section at the bottom of this page. There are several important issues with DR4 data to be aware of:

SkyMapper Detectors^ Back to top

The SkyMapper mosaic camera contains 32 CCDs of 4096 x 2048 pixels with a plate scale of ~0.50 arcsec/pixel. There are small gaps between the individual CCDs and the resulting field-of view is 2.37 deg x 2.39 deg. The mosaic fill factor is 91% of a 5.68 deg2 field-of-view. Each SkyMapper image is split into its 32 constituent CCDs, which are presented as separate files to the image cutout service.

SkyMapper Filters^ Back to top

The SkyMapper filter curves (with atmosphere) are shown below. We have tabulated colour transformations between SkyMapper and other standard filters, as well as predicted star colours and reddening corrections -- they can be found on the page here.

Note that the u-band filter (ultraviolet) is shortward of the Hydrogen Balmer break, while the v-band (violet) is placed between the Balmer break and the Ca H&K 4000AA-break. The u-filter also has a red leak, which needs an airmass-dependent correction. The two lines shown in the figure below are effective transmission curves including atmosphere for airmasses 1 and 2. The third (inset) panel shows the difference between the apparent u-magnitudes observed at airmass 1 and airmass 2 as a function of star colour. The redder the star the brighter it appears in u-band as it is observed closer to the horizon, because calibration stars get fainter, while the red leak keeps the flux bright. Additional information on the filter set can be found in Bessell et al. (2011).

World Coordinate System^ Back to top

Reduced SkyMapper images provided through this website and the SIAP service have been registered onto the sky using the TPV World Coordinate System (RA---TPV, DEC--TPV). TPV builds on the standard TAN projection by adding a general polynomial distortion suitable for wide-field cameras which is described in a set of additional PVi_m keywords. A typical TPV header is reproduced below:

Typical TPV FITS header
WCSAXES =                    2 / WCS dimensionality                             
CTYPE1  = 'RA---TPV'           / WCS projection type for this axis              
CTYPE2  = 'DEC--TPV'           / WCS projection type for this axis              
LONPOLE =                180.0 / WCS coordinate rotation: longitude             
LATPOLE =                  0.0 / WCS coordinate rotation: latitude              
CRVAL1  =        276.914401980 / RA of reference point                          
CRVAL2  =         -9.384011927 / DEC of reference point                         
CRPIX1  =         -2.900864031 / X reference pixel                              
CRPIX2  =        -63.233494932 / Y reference pixel                              
CUNIT1  = 'deg     '           / X pixel scale units                            
CUNIT2  = 'deg     '           / Y pixel scale units                            
CD1_1   =         -0.000138209 / WCS transformation matrix                      
CD1_2   =         -0.000000195 / WCS transformation matrix                      
CD2_1   =          0.000000225 / WCS transformation matrix                      
CD2_2   =         -0.000138246 / WCS transformation matrix                      
RADESYS = 'ICRS    '           / WCS reference frame                            
PV1_0   =          0.000042557 / WCS projection distortion parameter            
PV1_1   =          1.000461443 / WCS projection distortion parameter            
PV1_2   =          0.000190561 / WCS projection distortion parameter            
PV1_3   =         -0.000000012 / WCS projection distortion parameter            
PV1_4   =          0.001420356 / WCS projection distortion parameter            
PV1_5   =          0.000230202 / WCS projection distortion parameter            
PV1_6   =          0.000277876 / WCS projection distortion parameter            
PV1_7   =          0.000002261 / WCS projection distortion parameter            
PV1_8   =          0.000006422 / WCS projection distortion parameter            
PV1_9   =          0.000001831 / WCS projection distortion parameter            
PV1_10  =          0.000000521 / WCS projection distortion parameter            
PV1_11  =         -0.000000298 / WCS projection distortion parameter            
PV2_0   =          0.000071788 / WCS projection distortion parameter            
PV2_1   =          1.000381476 / WCS projection distortion parameter            
PV2_2   =          0.000493742 / WCS projection distortion parameter            
PV2_3   =          0.000000007 / WCS projection distortion parameter            
PV2_4   =          0.000440372 / WCS projection distortion parameter            
PV2_5   =          0.000927121 / WCS projection distortion parameter            
PV2_6   =          0.000817100 / WCS projection distortion parameter            
PV2_7   =         -0.000000338 / WCS projection distortion parameter            
PV2_8   =         -0.000001059 / WCS projection distortion parameter            
PV2_9   =         -0.000001035 / WCS projection distortion parameter            
PV2_10  =         -0.000004152 / WCS projection distortion parameter            
PV2_11  =         -0.000000226 / WCS projection distortion parameter             

As TPV is a relatively recent WCS parameterisation you may need to update older WCS libraries and software for best results. We have confirmed that TPV is supported by the following libraries and tools:

IRAF wcstools wcslib AST ds9 AstroPy Astromatic CDS Aladin IPAC Montage
2.16+ 3.8.4+ 5.0+ 5.7.3+ 7.0+ 1.1+ 3/2014 onwards 9+ 4.0+

Note: older versions of Montage (<=3.3) include wcstools v3.8.1 so this library must be updated to 3.8.4+ prior to compiling in order to support TPV. 

Astrometric Quality^ Back to top

The astrometry for DR4 is based upon the Gaia DR2 astrometric catalogue (Gaia Collaboration 2022). In a change from previous DRs, the WCS solutions were determined for each mosaic image, rather than CCD-by-CCD. The new solutions are more robust for sparse frames (especially shallow exposures in u and v), but tend to show systematic deviations in the extreme corners of the mosaic. Photometry from such regions has been prevented to contributing to the mean positions recorded in the dr4.master table. The median offset is 0.13 arcsec. The sky distribution of the median Gaia offsets per deg2, on a scale of 0 (white) to 0.5 arcsec (black) is shown below.

Photometric Quality^ Back to top

The new photometric zeropoint catalogue is anchored to Gaia low-resolution spectroscopy, which significantly improves the uniformity and precision of the calibration in u- and v-band compared to previous data releases.

The sky maps below show the median changes between DR4 and DR2 photometry.


Comparison with synthetic photometry of CALSPEC standard stars shows agreement across a range of colours that is typically better than 0.02 mag. The slight residual trends in u- and v-band may be due to small (1-5 nm) shifts in the filter bandpass relative to the expected filter curves.

The internal reproducibility of SkyMapper PSF photometry of isolated bright sources is on the order of 1%.


Known Issues | DR4^ Back to top

Click on the table headings to sort the table based on that value.

Issue ↕ First affected release ↕ Resolved in ↕ Last modified ↕
Extended-source photometry EDR Jul 13 2023, 19:26 AEST
Incorrect photometry when merging child objects DR1.1 DR4 Jul 13 2023, 19:25 AEST
Spatially varying PSF affecting photometry EDR Jun 05 2017, 15:55 AEST
ZIP file downloads in Safari (for info) EDR May 02 2016, 12:21 AEST