Beyond Cloud-9: The case for discovering more HI-rich failed halos

Abstract

HI-rich starless halos, should they exist, hold great promise for elucidating properties of dark matter halos. This Letter examines the properties of HI-rich failed halos at redshift zero across state-of-the-art cosmological simulations (FIREbox, NIVARIA-LG and Recal-EAGLE). First we compare two numerical analogs with Cloud-9, purported to be the first discovery of a starless HI-rich halo. We argue that differences may be driven by environmental factors, and/or the treatment of gas self-shielding -- which might further limit existing analytic schemes aimed at inferring dark matter halo information from 21 cm HI observations. We also find that the failed halo samples in the three simulations span different regions of the HI-gas-halo mass ($M_{\rm HI}-M_{\rm gas}-M_{\rm 200}$) plane. FIREbox objects occupy a very narrow regime, while NIVARIA-LG extends to a wider range of $M_{\rm 200}$ values - and achieves higher $M_{\rm HI}$ and $M_{\rm gas}$ values. Recal-EAGLE $M_{\rm HI}$ values are similar to FIREbox, albeit with lower gas and halo masses. Lastly, we predict that more HI-rich starless halos can be discovered by exploring the HI-poor regime in the local universe, rather than HI-rich populations at high redshift. Overall, we advocate for the allocation of resources to detect and characterize other HI-rich (and HI-poor) failed halos in the local universe, plus dedicated follow-up spectroscopic observations that scrutinize claims to the absence of a faint stellar component, and that assess their isolation status in detail.