In this work, we present the probabilities of mergers of binary black holes (BBHs) and binary neutron stars (BNSs) as functions of stellar mass, metallicity, specific star formation rate (sSFR), and age for galaxies with redshift z ≤ 0.1. Using the binary-star evolution (BSE) code and some fitting formulae, we construct a phenomenological model of cosmic gravitational wave (GW) merger events. By using the Bayesian analysis method and the observations from Advanced LIGO and Virgo, we obtain the relevant parameters of the phenomenological model (such as the maximum black hole mass is ). Combining the above model results with the galaxy catalog given by the EMERGE empirical galaxy model, we find the normalized probability of occurrence of a merger event varying with  for galaxies with z ≤ 0.1 is different from that in previous studies, that is, two peaks exist in this work while there is only one peak (log10(sSFR/yr −1) = −10) in the previous work. The sSFR value corresponding to the new peak is log10(sSFR/yr −1) = −12 and in line with the value () of NGC 4493, the host galaxy of BNS merger event GW170817. The new peak is caused by today's quenched galaxies, which give a large contribution to the total SFR at high redshift in the EMERGE empirical galaxy model. Moreover, we find that the BNS mergers are most likely detected in galaxies with age ∼11 Gyr, which is greater than previous results (6−8 Gyr) and close to the age of NGC 4993,  Gyr.