The attached plots compare REFL WFS 9 MHz and 45 MHz pitch and yaw signals between Hanford and Livingston. The calibration into watts is rough, and anyway the ground motion at the two sites is rather different, so direct comparison of the noise levels is not guaranteed to mean anything. However, both sites seem to be limited by PR3 damping noise at a wide range of frequencies, which in many cases dominates the rms of the total signal. There is also a contribution from HAM1 motion near and above 10 Hz.

For DOFs like PR3 angles, which are not under ASC control, one can do a coherence projection to ascertain their contribution to the REFL WFS error signals. The results are shown in the attached plots, with coherences projected into WFS A signals only. From poking at medm screens my understanding is that none of the Q phase signals are used for ASC. With those statements in mind...

• At both sites, most WFS signals are broadband coherent with PR3 local damping , especially at Hanford. PR3 motion significantly contributes to (or dominates) the rms of the WFS signals at 0.8 Hz for pitch or 1.2 Hz for yaw in many signals. Coherence of Hanford's WFS A RF9I P with the PR3 suspoint projections from the ISI is high below 1 Hz but low above 1 Hz, suggesting that the motion above 1 Hz is mostly sensor noise from the local damping. Tests at Livingston previously showed that PR3 motion induces beam spot motions on the test masses that are a fair fraction of 1 mm (LLO:58623).
• Hanford's WFS signals are limited by IM motion from ~5 to 30 Hz, as seen previously (LHO:63616). This noise is not seen in Livingston because the local damping cutoffs are more aggressive (LLO:60551).
• At both sites HAM1 motion still contributes from ~5 Hz to 30 Hz. Given the above bullet, feedforward subtraction of HAM1 motion will probably be of limited effectiveness at Hanford until the IM damping noise is reduced.

At higher frequencies the signals are limited by shot noise and dark noise (see, e.g., LHO:62976 from before the HAM1 power reduction). At other frequencies where there is no dominant coherence term plotted, the WFS signals at Hanford are coherent with optics under ASC control and therefore some actual budgeting will be required. The 0.8 Hz and 1.2 Hz peaks in PR3 motion are also seen in the oplevs, so some narrowband oplev damping may be an easy way to see if reduction in PR3-dominated beam-spot motion gives any noise improvement.