We began looking into whether the recent noise made by GV-20 is new or has been increasing with time by making spectrograms of the period of the audio recording in alog 91930. We optimized the 30s spectrograms to make the 4 events in the audio recording as visible as possible and then we made 600 second spectrograms of openings and closings over the past few years. These 600s spectrograms are about twice as long as the gate valve opening or closing so that we could compare the time while it was moving to the time before and after. In cases where the sound and vibration level at EX was not much louder than during the Sept. 8 events, we found similar densities of glitches during openings and/or closings back to 2020, and much lower densities of glitches (from other sources) when the valves were not actuating (see figure).

Miranda, Alex, Robert