Based on a dynamical formation model of a supermassive black hole ( SMBH ) , we estimate the expected observational profile of gravitational wave at ground-based detectors , such as KAGRA or advanced LIGO/VIRGO .
Noting that the second generation of detectors have enough sensitivity from 10 Hz and up ( especially with KAGRA owing to its location at less seismic noise ) , we are able to detect the ring-down gravitational wave of a BH with the mass M < 2 \times 10 ^ { 3 } M _ { \odot } .
This enables us to check the sequence of BH mergers to SMBHs via intermediate-mass BHs .
We estimate the number density of galaxies from the halo formation model and estimate the number of BH mergers from the giant molecular cloud model assuming hierarchical growth of merged cores .
At the designed KAGRA ( and/or advanced LIGO/VIRGO ) , we find that the BH merger of its total mass M \sim 60 M _ { \odot } is at the peak of the expected mass distribution .
With its signal-to-noise ratio \rho = 10 ( 30 ) , we estimate the event rate R \sim 200 ( 20 ) per year in the most optimistic case , and we also find that BH mergers in the range M < 150 M _ { \odot } are R > 1 per year for \rho = 10 .
Thus , if we observe a BH with more than 100 M _ { \odot } in future gravitational-wave observations , our model naturally explains its source . Published as Astrophysical Journal 835 ( 2017 ) , 276 . \url http : //dx.doi.org/10.3847/1538-4357/835/2/276 arXiv:1610.09505v3 .