Shear measurement

Shear object detections and measurements in four “counterfactual” deep coadd images with different shear steps, plus a fifth with no shear. All shear measurements are stored in the catalog Data products.

Counterfactual images

The GalSim Shear class is used to create five “counterfactual” versions of the deep coadd images with different artificial linear shears introduced, using the “reduced shear” option with a step value (g1) of 0.01 (GalSim documentation; Rowe et al. 2025). The application of shear to an image essentially performs a convolution that adds ellipticity with a set minor-to-major axis ratio and position angle, but conserves area.

The five counterfactual images and their shear directions are:

• ns: no shear

• 1p: x-direction

• 1m: y-direction

• 2p: 45 degree diagonal

• 2m: 135 degree diagonal

Detection and measurement

All objects detected with a significance of \(> 5\sigma\) in a given counterfactual image have their shapes measured.

Two approaches are used for shape measurement: 1. A single 2D gaussian is fit jointly across the r, i, and z-band deep coadd images. The resulting ellipticity components in the image x-y (gauss_g1) and diagonal (gauss_g2) directions are measured. 2. The 2D gaussian-weighted moments on single band (r, i, z) images are fit separately. The resulting ellipticity components of the sum, weighted by inverse-variance, in the image x-y (pgauss_g1) and diagonal (pgauss_g2) directions are measured.

In addition to the shape measurements above, the positions, fluxes, gaussian moments, and PSF shapes are also measured.

Resulting catalog

The results of shear detection and measurement are stored in the ShearObject catalog.

If a given galaxy is detected with \(> 5\sigma\) in all five counterfactual images, then it will populate five rows of the ShearObject table (with five separate shearObjectId).

The process of detecting and measuring objects in the counterfactual images, and the resulting ShearObject table, is completely independent of the process that creates the Object table. These two tables are not joinable and have independent (unique and unrelated) object identifiers.