- triggered BAT twice (trigger# 954519, 954520). Event data are merged to trigger954519. T0 refers to T0 for trigger954519. Also, since HEASARC data has a larger data gap (from T+63 s to T+339), we decided to use the pass data (from the BAT team machine) instead, which only have a gap from T+63 s to T+160 s. The data gap is dued to a problem in saving data because of the double triggers (GCN Circ. 27012).
- using the attitude file (the *sat* file) from the BAT team machine. This is because the HEASARC attitude file doesn't cover time after ~ T+1000 s, and make incorrect mask-weighted counts in that time range, which causes incorrect duration estimation. 
- Obvious data gap within the burst duration.