With the purpose to shed some more light on the HWS measurement, we have analyzed the raw images before and after the appearance of the feature in the upper part of the wavefront.

Figure 1 shows the difference between the two raw images. As it can be seen, there is a significant modification of the illumination pattern on the HWS as soon as the ITF has unlocked. The change in the illumination pattern is responsible for the appearance of the feature.

This change in the illumination pattern can be explained with a displacement of the HWS probe beam (due to a misalignment of one of the optics or benches along the path) or with some spurious other beam (1064 nm or 532 nm) reaching the HWS plate. With the available data, it is not possible to discriminate between the two effects, even if a check of the LC signals (figure 2) seems to exclude the displacement of the probe beam.

To investigate the hypothesis that the structure in the upper part of the HWS map is caused by the green beam, Diego modified the automation configuration to keep it OFF after the unlock. We then performed a new measurement with HWS-DET. For this purpose, I switched OFF the HWS-INJ SLED and switched ON the HWS-DET SLED.

The ITF was locked at CARM NULL 1F at 11:04 UTC. Acquisition started at 14:28 UTC (folder 20251029T1528), and the ITF gently unlocked at 14:36 UTC (WF14).

The structure was also present in the maps (see figure).
The measurement was interrupted, and the lock acquisition was restarted.

The issue with the HWS intensity changing during the gentle unlock is a long standing story, I found several entries reporting it and discussing the solution (60377, 60362, 60344, 59433, 59434, 59347).

In one of them (59433), the issue seems to be traced back to the ETMs misalignment during the unlock (but keeping them aligned seems worse).

In any case, the only solution seems to be the re-processing of the data. 

In the meantime, we performed some additional data analysis to further investigate the observed effects. The following results summarize the outcomes and comparisons made with previous datasets:

• First, I verified that a beam shift occurs only on HWS-DET during the unlock, and not on HWS-INJ.
  By taking the difference between the CCD illumination patterns before and after the unlock of the last measurement 20251029T1009, no misalignment or spurious beam was observed (see Fig. 1). 

• I reprocessed the HWS-DET data from acquisition (2) – 20251028T1829, reported in entry 68051, as suggested by Alessio.
  The reprocessing took WF21, the first wavefront acquired after the unlock, as the reference.  Some WFs acquired after the unlock are shown in Fig. 2.
  Then, I extracted the curvature and compared it with the one obtained from the last measurement performed with DET (20230316T0008).
  The two results are consistent with each other, as shown in Fig. 3. The current curvature is slightly lower, possibly due to a slightly lower input and/or circulating power in the arm, or because some curvature is lost as the YAG beam effect is at the right edge of the map (i.e. the SLED beam is decentered on the NI mirror).

• I revisited the last absorption measurement performed with HWS-DET (20230316T0008), which had been reprocessed afterwards.
  Before reprocessing, considering for example the WF9, corresponding to approximately 30 s after the unlock, it appeared as shown on the right side of Fig. 4. On the left side of the same figure, the difference between the illumination profile of the reference acquired before the unlock and that acquired after the unlock is shown. Thus, comparing this difference map with the one reported in the entry  68056,  in 2023 the problem of beam displacement was more evident. Without reprocessing the wavefronts, the YAG never appears in the map (see an example is WF59- Fig. 5).
  By reprocessing the images with respect to WF8, the YAG effect becomes visible in WF9, as shown in Fig. 6, evolving over time in a manner similar to that observed in the most recent dataset.