TJ Massinger, Jenne Driggers, Derek Davis

Summary: Subtraction of technical noise using DHARD/CHARD and MICH/PRCL/SRCL control yields ~4% increase in inspiral range (assuming the calibration is correct)

Jenne requested that the noise subtraction pipeline be used to estimate how much sensitivity could be gained from subtracting ASC/LSC control noise. The suggested noise sources to target were derived from a Bruco run that Sheila generated.

I ran the offline noise subtraction pipeline that was used for production jitter subtraction in O2 on two channel groups:

ASC: 

ASC-DHARD_P_OUT_DQ

ASC-DHARD_Y_OUT_DQ

ASC-CHARD_P_OUT_DQ

ASC-CHARD_Y_OUT_DQ

LSC (labeled "vertex" in subtraction result page):

LSC-MICH_OUT_DQ

LSC-PRCL_OUT_DQ

LSC-SRCL_OUT_DQ

The measured transfer functions can be seen in the subtraction results page in section 3.1 for ASC and 3.2 for LSC ("vertex").

The attached figure shows an ASD (8 second FFT length) before and after subtraction zoomed into the impacted region of 5-200 Hz.

The inspiral range calculated with a minimum frequency of 10 Hz increased from 77.3 Mpc to 80.4 Mpc (4%). 

Caveat: Calibration is not final, these are preliminary values for inspiral range.