I'm attaching a screenshot showing a comparison of Camilla's broadband calibration measurement (LHO 70854) with the hot OM2 (LHO 70849) vs a measurement taken last week while the OM2 was still cold. The pink trace shows a measurement taken last week on 6/21 and the green was taken while the OM2 was heated. 

I want to point out the right column, which shows H1:GDS-CALIB_STRAIN_CLEAN / H1:CAL-PCALY_RX_PD_OUT_DQ (since Nonsens cleaning is currently turned off as per LHO 70654 H1:GDS-CALIB_STRAIN_CLEAN is the same as H1:GDS-CALIB_STRAIN_NOLINES). The OM2 TSAMS changes have resulted is roughly 2% additional error at 20Hz, among several other changes that I can't fully explain yet. It's worth noting here that the GDS pipeline applies time dependent corrections to H1:CAL-DELTAL_EXTERNAL in real time as picked up by the TDCFs ("the kappas"). The fact that there are such noticeable differences in H1:GDS-CALIB_STRAIN_CLEAN / H1:CAL-PCALY_RX_PD_OUT_DQ indicates that the effects due to the OM2 being heated on the sensing function are not being tracked by the Kappas. If they were then we'd see two overlaid traces on the PCALY to GDS-CALIB_STRAIN plot.

I'm also leaving a screenshot showing how the kappas changed after the OM2 adjustment (kappas_scope.png). KappaC and Fcc changed by ~1.6% and 3%, respectively. Notably, the UIM kappa was about 2% off and then slid to ~1? 

I'm also attaching trends of the systematic error lines (sys_err_lines_scope.png) from /ligo/home/jeffrey.kissel/Templates/NDScope/uncertainty.yaml (not a typo). Several of the systematic error lines experience a drift after the OM2 TSAMS change. Jeff suggested that changes to the sensing function, which will impact the systematic error measurements (1/(1+G)), could "bleed" into and influence the line measurements that the kappa calculations are based on. More on this later.

refs in attached dtt screenshot:

  refs 0-9: same as listed in LHO 70705

  6/27 om2 hot:
    ref 10: deltal_external
    ref 11: gds calib strain clean