Speaker: Miranda Gorsuch, Physics PhD Graduate Student
Abstract: Weak gravitational lensing encodes information of the structure of matter as well as the expansion history of the observable universe. However, the weak lensing signal is small, and requires measurements of the signal across a large sample of galaxies. In the era of the Vera C. Rubin Observatory and the Legacy Survey of Space and Time, billions of shape measurements will be detected across 18,000 square degrees. A galaxy sample of this scale affirms that we are in the era of precision cosmology; now considerable care must be taken to accurately measure the shapes of galaxies in order to preserve the weak lensing signal against systematics. Part of the work for maintaining accurate shape measurements includes a recent method for stacking images - or coadding - referred to as cell-based coadds. This coaddition scheme better preserves the shape measurements of galaxies by neglecting input images that do not fully cover the region of the coadd, which if included introduce systematic biases in the shape measurements. I will present my current work that focuses on interpolated pixels within cell-based coadd images. Allowing images with a large fraction of masked pixels into the final coadd may impact shear measurements; on the other hand, discarding too many input images may affect final coadd depth and object detection. I have developed a pipeline to determine the average fraction of input images lost due to the images going over the allowed fraction of masked pixels. The next stage of this project will use the metadetection algorithm to measure the shear response on simulated image data. The purpose of this stage is to characterize how shear measurements are impacted by changes in the maximum mask fraction of input images. Finally, an object detection stage will be done to understand impacts in the variation of coadded image depths.