A novel Cosmic Filament catalogue from SDSS data.

In this work we present a new catalogue of Cosmic Filaments obtained from the latest Sloan Digital Sky Survey (SDSS) public data. In order to detect filaments, we implement a version of the Subspace-Constrained Mean-Shift algorithm, boosted by Machine Learning techniques. This allows us to detect cosmic filaments as one-dimensional maxima in the galaxy density distribution. Our filament catalogue uses the cosmological sample of SDSS, including Data Release 16, so it inherits its sky footprint (aside from small border effects) and redshift coverage. In particular, this means that, taking advantage of the quasar sample, our filament reconstruction covers redshifts up to $z=2.2$, making it one of the deepest filament reconstructions to our knowledge. We follow a tomographic approach and slice the galaxy data in 269 shells at different redshift. The reconstruction algorithm is applied to 2D spherical maps. The catalogue provides the position and uncertainty of each detection for each redshift slice. We assess the quality of the detections with several metrics, which show improvement with respect to previous public catalogues obtained with similar methods. We also detect a highly significant correlation between our filament catalogue and galaxy cluster catalogues built from microwave observations of the Planck Satellite and the Atacama Cosmology Telescope.