Oshino-san, thank you very much for modification!

I attach the plots to show the significant difference which starts from ~4:00 UTC.

Although the number of trigger was redulced by clustering, the SNR was not changed. It is clear to see the central frequency of glitches (left top of Fig.1).