Ultra Fast Image Generator (UFig) BCC

Authors: Lukas Gamper, Claudio Bruderer, Michael Busha, Joel Bergé, Tomasz Kacprzak, Alexandre Refregier, Adam Amara

Using our developed Ultra Fast Image Generator (UFIG; Bergé et al. (2013)) we simulated a total 136 coadd tiles, corresponding to an area of about 68 square degrees. Our image simulations are based on the BCC simulations (Blind Cosmology Challenge; Busha et al. (2013)) for the Dark Energy Survey (DES). We chose to simulate the full SPT-E field. As the star density is unusually high in some parts of this region in the BCC catalogs, we selected an equivalent region 10 degrees farther North.

Simulating one tile on a single core took less than 2 minutes. Creating the SExtractor catalogs took another 2 minutes. Thus, we were able to simulate this large area in less than 5 minutes by running these processes in parallel.

This release is a first version and is thought to be used for testing purposes only.

Link to the simulated coadd tiles

Images

We calibrated our simulations using the technical properties of the DECam such that the images resemble DES coadd-images as closely as possible. The saturation is set to approximately 10 times the values of single exposures. We found that this fits the coadd data better than the values reported in the DES coadd tile-headers. We simulated only i-band coadds.

We used a circular Moffat profile without spatial variation as a model for the Point Spread Function. Our target profile has a Beta of 3.5 and a seeing FWHM of 0.9 arcseconds as representative for DES.

BCC catalogs

We used the following catalogs to generate our images:

Aardvark_v0.5d_des_masked
Aardvark_0.5c_stars

SExtractor settings

We set the detection threshold to 1.7 sigmas for at least 5 contiguous pixels. The seeing is set by the PSF_FWHM of the target Moffat profile specified above.

MEDS

Multi Epoch Data Structures (MEDS) is the file format used in the DES weak lensing community. Objects found in the coadded images are extracted and bundled with the corresponding single exposure image of that object. This allows measuring the shape using information contained in single exposures. Further information can be found here: https://cdcvs.fnal.gov/redmine/projects/deswlwg/wiki/Multi_Epoch_Data_Structure

As we only generated coadded images, the contained single exposure is a copy of the coadded image. Additionally, we used a constant weight map to create the MEDS files.

.fits: Coadd image
_matching.cat: Catalog matching SExtractor catalog to truth catalogs
.sexcat: SExtractor catalog
_bkg.fits: SExtractor background map
_seg.fits: SExtractor segmentation map
.meds.fits: MEDS file containing images and maps
.meds.cat: MEDS file containing objects ASCII-catalog

The _matching.cat ASCII-catalog has the following columns:

[Object number in .sexcat] [class/star-parameter] [Number of the according truth catalog] [Object ID in the truth catalog]

If no match is found, the entry is set to -1.

References