Jointly setting upper limits on multiple components of an anisotropic stochastic gravitational-wave background

With the increasing sensitivities of the gravitational wave (GW) detectors and more detectors joining the international network, the chances of detection of a stochastic GW background (SGWB) are progressively increasing. Different astrophysical and cosmological processes are likely to give rise to backgrounds with distinct spectral signatures and distributions on the sky. The observed background will therefore be a superposition of these components. Hence, one of the first questions that will come up after the first detection of a SGWB will likely be about identifying the dominant components and their distributions on the sky. Both these questions were addressed separately in the literature, namely, how to separate components of isotropic backgrounds and how to probe the anisotropy of a single component. Here, we address the question of how to separate distinct anisotropic backgrounds with (sufficiently) different spectral shapes. We first obtain the combined Fisher information matrix from folded data using an efficient analysis pipeline pystoch, which incorporates covariances between pixels and spectral indices. This is necessary for estimating the detection statistic and setting upper limits. However, based on a recent study, we ignore the pixel-to-pixel noise covariance that does not have a significant effect on the results at the present sensitivity levels of the detectors. We establish the validity of our formalism using injection studies. We show that the joint analysis accurately separates and estimates backgrounds with different spectral shapes and different sky distributions with no major bias. This does come at the cost of increased variance. Thus making the joint upper limits safer, though less strict than the individual analysis. We finally set joint upper limits on the multicomponent anisotropic background using Advanced LIGO data taken up to the first half of the third observing run.