Title: Gravitational wave memory of the binary black hole events in GWTC-2

Abstract: Gravitational wave (GW) memory is an important prediction of general relativity. Existing works on the GW memory detection focus on the waveform analysis. It is hard for waveform analysis method to detect the GW memory due to its quasi-direct current behavior and weakness. We implement a completely different scheme in this work to estimate the GW memory. In this scheme, we firstly apply the Bondi-Metzner-Sachs method to calculate the GW memory of binary black hole based on numerical relativity simulation. Then we construct a surrogate model to relate binary black hole's parameters and the GW memory. Afterwards we apply this surrogate model together with Bayesian techniques to estimate the GW memory of the 48 binary black hole events recorded in GWTC-2. The GW memory corresponding to the all 48 events has been estimated. The most interesting results are for GW190814. The corresponding GW memory is about $-1\times10^{-23}$ and $1\times10^{-23}$ for Hanford detector and Livingston detector respectively. At the same time we find with 3$\sigma$ C.L. that the memory strain of GW190814 is negative on Hanford detector while positive on Livingston detector.