The image modeling framework "scarlet" (Melchior et al. 2018) enables the joint use of multiple data sets of the same celestial scene. It can be used to improve source modeling and source separation by working on multi-band images and/or by combining low-resolution with high-resolution images. It is specifically developed for joint pixel-level processing of data from LSST, WFIRST, and Euclid. I will introduce the method, discuss the statistical challenges in interpreting scenes consistently across multiple data sets, and demonstrate the improvement for precision photometry and shape measurements that are achievable with scarlet.
42nd COSPAR Scientific Assembly
- Pub Date:
- July 2018