The SMASH Survey


The Survey of the Magellanic Stellar History (SMASH) is using DECam to map 480 square degrees of sky to depths of ugriz~24 with the goal of identifying broadly distributed, low surface brightness stellar populations associated with the stellar halos and tidal debris of the Magellanic Clouds. It will eventually contain measurements of approximately 250 million objects distributed in discrete fields spanning an area of about 2400 square degrees. The first data release (DR1) contains ~100 million objects from 61 observed fields. Browse these pages to learn more about SMASH and to access the data. The SMASH overview paper (Nidever et al. 2017) describes the survey in detail, including its goals, survey strategy, reduction, and calibration.

SMASH at a Glance
Area covered480 deg2 spanning ~2400 deg2
Depth (5σ, ugriz)23.9, 24.8, 24.5, 24.2, 23.5 mag
Seeing (ugriz)1.22, 1.13, 1.01, 0.95, 0.90 arcsec
Number of fields197
      -   DR161
Number of DECam exposures5,809
      -   DR12,480
Number of objects~420,000,000
      -   DR1101,425,210
Number of measurements~4,000,000,000
      -   DR1722,653,189
Photometric precision~1% in u and 0.5-0.7% in griz
Photometric calibraton accuracy~1.3% in all bands
Astrometric accuracy~20 mas
Back to Top


Using old main sequence stars as tracers, SMASH will reveal the relics of the formation and past interactions of the Clouds down to surface brightnesses equivalent to Σg=35 mags arcsec2. The main project goals are:
  • Search for the stellar component of the Magellanic Stream and Leading Arm. The detection of stellar debris in these structures would make them the only tidal streams with known gaseous and stellar components in the Local Group. This would not only be invaluable for understanding the history and observable consequences of the Magellanic interaction, but would give us a dynamical tracer of the MW's dark halo and a way to probe the MW's hot halo gas via ram pressure effects.
  • Detect and map the smooth components of the Clouds, including their extended disks and potential stellar halos. The size of the LMC's stellar disk is a direct probe of the tidal radius of the LMC, with which we can explore the dark matter halos of the LMC and MW.
  • Detect and map potential streams and substructure in the Magellanic periphery not associated with HI features. These would probe stages in the formation and interaction of the Clouds at times earlier than the HI dissipation timescale.
  • Derive spatially resolved, precise star formation histories covering all ages of the MCs and to large radii, thus providing detailed information on their complete evolution.
  • Enable many community-led projects, including studies involving the LMC/SMC main bodies, Galactic structure, discovery of variable objects, and background galaxy populations.

Back to Top

First Data Release

The figure below shows the 197 SMASH fields. Sixty one of the fully-calibrated deep fields are being publicly released in the first SMASH data release (DR1) which contains ~700 million measurements of ~100 million objects in fields sampling the ~2400 deg2 region of the SMASH survey (blue hexagons in the figure below). The main data access is through this prototype version of the NOAO Data Lab. Access and exploration tools include a custom Data Discovery tool, database access to the catalog (via direct query or TAP service), an image cutout service, and a Jupyter notebook server with example notebooks for exploratory analysis.

Catalogs and FITS images: Images, intermediate data products, and final catalogs (in FITS binary formats) are also available through the NOAO High Level Data Products FTP site. The raw images as well as the NOAO Community Pipeline (CP) reduced InstCal, Resampled and single-band Stacked image are available in raw/, instcal/, resampled/, and stacked/ directories, respectively, and grouped in nightly subdirectories (see here for more information on the NOAO archive's file naming convention). Each subdirectory has an "inventory" file that gives useful information about each fpack-compressed FITS image (e.g., exposure number, field name, time stamp, filter, exposure time). The PHOTRED-ready FITS files and other associated files (PSF, photometry catalogs, logs, etc.) as well as the multi-band stacks are available in the photred/ directory. The final binary FITS catalogs are in the catalogs/ directory with seven files per field:
  • FIELD_exposures.fits.gz - Information on each exposure.
  • FIELD_chips.fits.gz - Information on each chip.
  • FIELD_allsrc.fits.gz - All of the individual source measurements for this field.
  • FIELD_allobj.fits.gz - Average values for each unique object.
  • FIELD_allobj_bright.fits.gz - Bright stars from allobj used for cross-matching between fields.
  • FIELD_allobj_xmatch.fits.gz - Crossmatches between SMASH and Gaia, 2MASS and ALLWISE.
  • FIELD_expmap.fits.gz - The "exposure" map per band.
This file gives more detailed descriptions of the columns in these catalogs.

Database: There are six tables in the database that were populated (slightly modified) using the FITS catalogs mentioned above: field, exposure, chip, source, object, and xmatch. The "field" table includes summary information for each field. There are also two database "views" available, which are filtered versions of the object table. The "stars" view contains objects with abs(sharp)<0.7 and ndet>1, while the "galaxies" view contains objects with abs(sharp)>0.7, ndet>1, and prob<0.2. This page gives information on the database schema and columns. Back to Top

Data Reduction and Calibration

The SMASH data reduction of the DECam data makes use of three separate software packages: (1) the Community Pipeline for instrument signature removal, (2) PHOTRED for PSF photometry, and (3) SMASHRED, custom software written for PHOTRED pre- and post-processing of the SMASH data.

Instrument Signature Removal: The NOAO DECam Community Pipeline (Valdes et al. 2014) was jointly developed by the Dark Energy Survey Data Management (DESDM) team and NOAO to produce reduced images for the community. The CP performs bias and crosstalk correction, masking of bad pixels and other artifacts, linearity correction, flat field and illumination calibration, fringe pattern subtraction, astrometric and photometric calibration, sky pattern subtraction, remapping and coaddition.

PSF Photometry: PHOTRED is an automated and robust PSF photometry pipeline based on the DAOPHOT suite of programs (Stetson 1987, 1994). It performs WCS fitting, single-image PSF photometry (ALLSTAR), source matching across multiple images, forced PSF photometry across multiple exposures using a master source list created from iterative detection from a deep multi-band "detection" stack (ALLFRAME), aperture corrections, and dereddening. PHOTRED was run on each nightly separately after the CP-reduced images were pre-processed using the "SMASHRED_PREP.PRO" script. The precision of the photometry for bright stars is ~1% in u and 0.5-0.7% in griz.

Calibration: Since most of the SMASH data are in disconnected "island" fields and little prior ugriz data exists in the southern sky, we used traditional techniques of calibrating our data with observations of standard star fields (on photometric nights; SDSS equatorial fields) and extra calibration exposures (for non-photometric nights) from the CTIO 0.9m telescope. The DECam standard star field exposures were processed with STDRED, a sister package to PHOTRED, which works in a similar manner. Then, custom software (SOLVE_TRANSPHOT.PRO) was used to derive the photometric transformation equations for all of the DECam standard star field data combined. Finally, the DECam science data were calibrated using custom SMASH software (SMASHRED_CALIBRATE_FIELD.PRO) that ties all chip data for a given field onto the same photometric zero-point (via overlaps) using an ubercal technique and calibrates the zero-point using the DECam transformation equations for photometric nights or the 0.9m data for non-photomeric nights. The accuracy of the photometric calibration is ~1.3% in all bands.

Back to Top