We combine information from the clustering of HI galaxies in the 100 % data release of the Arecibo Legacy Fast ALFA survey ( ALFALFA ) , and from the HI content of optically-selected galaxy groups found in the Sloan Digital Sky Survey ( SDSS ) to constrain the relation between halo mass M _ { h } and its average total HI mass content M _ { HI } .
We model the abundance and clustering of neutral hydrogen through a halo-model-based approach , parametrizing the M _ { HI } ( M _ { h } ) relation as a power law with an exponential mass cutoff .
To break the degeneracy between the amplitude and low-mass cutoff of the M _ { HI } ( M _ { h } ) relation , we also include a recent measurement of the cosmic HI abundance from the \alpha .100 sample .
We find that all datasets are consistent with a power-law index \alpha = 0.44 \pm 0.08 and a cutoff halo mass \log _ { 10 } M _ { min } / ( h ^ { -1 } M _ { \odot } ) = 11.27 ^ { +0.24 } _ { -0.30 } .
We compare these results with predictions from state-of-the-art magneto-hydrodynamical simulations , and find both to be in good qualitative agreement , although the data favours a significantly larger cutoff mass that is consistent with the higher cosmic HI abundance found in simulations .
Both data and simulations seem to predict a similar value for the HI bias ( b _ { HI } = 0.875 \pm 0.022 ) and shot-noise power ( P _ { SN } = 92 ^ { +20 } _ { -18 } [ h ^ { -1 } { Mpc } ] ^ { 3 } ) at redshift z = 0 .