J. Kissel

I've processed actuation function measurements from O3 thus far under nominal IFO conditions (namely, the nominal spot positions on test masses). 

In short: each ETMX actuator stage strength has not changed one bit, and our 2019-04-16 model, informed by measurements taken on 2019-04-03 (see processing of that data in LHO aLOG 48378) is still valid, in the frequencies where it matters.

Attached are plots confirming this conclusion. The "model" in the plots is the 2019-04-16 model,
    /ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/params/modelparams_H1_20190416.py
For each actuation stage of ETMX, UIM, PUM, and TST, I show three plots:
     Page 1: the model compared against measurement, 
     Page 2: the ratio of model compared against measurement, and 
     Page 3: the resulting GPR fit to the residuals.

In the last page, we see that -- in the frequency region which matters for each stage -- the uncertainty estimate (shaded blue) on the systematic error (solid blue line), informed by the collection of data (blue dots) are consistent with unity magnitude and zero phase, i.e. there is no systematic error. Outside the region where it matters, there is systematic error, and we know that these are flaws in the dynamical model of the ETMX dynamics used. These effects are known, (see detailed description in LHO aLOG 47331), but because they are in frequency regions that do not matter, we've not and will not put energy in to fixing them.

This allows us to comfortably update the GPR posterior distributions of unknown systematic error on the actuator stages, which in turn inform the response function uncertainty, improving our estimate of the overall unknown systematic error from these actuation model parameters.

These plots are created by the following script:
    /ligo/svncommon/CalSVN/aligocalibration/trunk/Runs/O3/H1/Scripts/FullIFOActuationTFs/process_actuationmeas_collection_20190716.py

I have not yet produced any new HDF5 files, nor updated the uncertainty budget. I need to talk to Lilli before understanding how to do so.

Also still to come -- analysis comparing the ETMX actuation strength as measured by an IFO with its nominal spot positions (i.e. the L2 and L3 data from all dates included in this entry) and with an IFO whose ETMY spot position has been moved close to center (i.e. L2 and L3 data from LHO aLOG 50382).