This is not a quantitative argument or anything, but just an observation.

In the morning yesterday, I noticed that the beam shape at AS AIR looked ugly presumably baecause of a non-optimum alignment servo configurations which seemed to have tuned the alignment where noise below 100 Hz was more non-stationary. After the commissioning work last night (alog 20699), the interferometer alignment seems to be better and indeed the beam in the AS AIR camera looks more Gaussian. The attached are a camera image of AS AIR from yesterday and this after noon. Both images are from the time when the interferometer was locked on DC readout at 24 W with the ASC loops fully engaged. The beam diverter happened to be open for both images.

As shown in these picutures, the beam of yesterday has some halo structure (mostly shown in purple) surrounding the bulk of the light while the one from this after noon does not have such a dirty halo. Interestingly, the white saturation area at the center in today's image appeaeres to be smaller than the one from yesterday. This can be an indication of less higher order modes (including misalignment) in the current alignment configuration. I checked the exposure time of two images and confirmed that were set to the same exposure time, 1800 usec.

Also, in yesterday's morning, I saw a ring-looking halo component moving up and down in the AS AIR camera behind the bulk of the light on a time scale of a few seconds. This ring-looking component can be actually seen in the above image -- it is on the upper left from the center Gaussin component and seen in white/purple. I did not get a chance to identify what optic was making the halo move up an down.