We quantify the frequency of companions of low redshift ( 0.013 < z < 0.0252 ) , dwarf galaxies ( 2 \times 10 ^ { 8 } M _ { \odot } < M _ { star } < 5 \times 10 ^ { 9 } M _ { \odot } ) that are isolated from more massive galaxies in SDSS and compare against cosmological expectations using mock observations of the Illustris simulation .
Dwarf multiples are defined as 2 or more dwarfs that have angular separations > 55 \arcsec , projected separations r _ { p } < 150 kpc and relative line-of-sight velocities \Delta V _ { LOS } < 150 km/s .
While the mock catalogs predict a factor of 2 more isolated dwarfs than observed in SDSS , the mean number of observed companions per dwarf is N _ { c } \sim 0.04 , in good agreement with Illustris when accounting for SDSS sensitivity limits .
Removing these limits in the mock catalogs predicts N _ { c } \sim 0.06 for future surveys ( LSST , DESI ) , which will be complete to M _ { star } = 2 \times 10 ^ { 8 } M _ { \odot } .
The 3D separations of mock dwarf multiples reveal a contamination fraction of \sim 40 % in observations from projection effects .
Most isolated multiples are pairs ; triples are rare and it is cosmologically improbable that bound groups of dwarfs with more than 3 members exist within the parameter range probed in this study .
We find that < 1 % of LMC-analogs in the field have an SMC-analog companion .
The fraction of dwarf ‘ ‘ Major Pairs ’ ’ ( stellar mass ratio > 1:4 ) steadily increases with decreasing Primary stellar mass , whereas the cosmological ‘ ‘ Major Merger rate ’ ’ ( per Gyr ) has the opposite behaviour .
We conclude that cosmological simulations can be reliably used to constrain the fraction of dwarf mergers across cosmic time .