Files
aimsight-labeling/auto/analyses/NORDEND_2026-10-13T0745_2026-10-24T1630/report.md
T
maximeandClaude Opus 5.5 ddb17c2189 auto: labels requalified under the 5-noise rules (only what stands out from the noise): 32 analyses refined, the 6 paused periods verified, the long labels requalified with the fixes the period analysts found; 1153 -> 961 labels
- 32 analyses: each label kept, rewritten or removed by its analyst (section Requalification of the report; mm magnitudes where the targets have them)
- SNCF01 x3, WILDSTRUBEL 2026-09-01 -> 2026-10-06 x3: verified from their existing labels under the new rules
- long labels (no analysis): auto/analyses/long-labels/<STATION>/; NORDEND-169 and -195 run to 07/11, so NORDEND-283 and -284 (their continuation) are removed
- reading-rules.md, stations.md: the citations of removed or rewritten labels point to their successor, or to [expert]

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-17 23:58:00 +01:00

8.1 KiB

NORDEND: labels requalified, 13 to 24 October

You asked me to requalify the labels of 13 October 07:45 → 24 October 16:30 (UTC) under the new rules: only what stands out from the noise of the corrected curves.

In short

35 labels remain out of 44. The 9 removed labels described missing points or a camera movement that the correction absorbs. Don't trust the curves during the labelled moments: the camera cycle on most days, the dawn mist of the 14th, the evening of the 16th, the dim-light jumps of the 21st and 22nd, and the rain of the 23rd. I see no real movement of the structure. Since the rain of the 23rd, the curves sit a little apart from where they were (the reference base changed).

What happened

The daily camera cycle (confirmed). Each day, the camera turns a little as its housing warms, then comes back in the evening. The correction does not remove it all. PR18 dips the most, by 3 to 5 mm on the sunny days (14th, 15th, 16th, 19th). On the overcast days of the 21st, 22nd and 24th the remainder is small, 1 to 3 mm, on a few targets in the vertical. It still stays above the noise, so these days keep their label. I removed only the cooling of the 23rd evening: the correction absorbs it completely. The targets are 84 to 150 m from the camera; the rock face does not move with the temperature.

Faint targets in dim light, and the dusk (likely). When the light falls (overcast midday, dusk), the spots weaken and the curves of their neighbours jump. PR05, PR14, PR15 and PR16 jump by 2 to 13 mm, PR15 by 9 mm at dusk on the 21st. These labels stay. The labels of the night rejections of the faint targets on the left edge (PR09, CTR01, PR04, PR07) are removed. Their curves are missing, but losing them does not move the other targets. None of them is a reference of the correction.

References disturbed (confirmed). Each time one of the three references is badly seen, all the targets jump together. On the evening of the 16th, CTR04 weakens and the targets jump by up to 10 to 14 mm. On the afternoon of the 14th, the sun lights the rock behind CTR02 and CTR03. CTR03 is then seen less well, and PR11 to PR13 drop by about 1 mm for half an hour. I kept that label, but it now describes this effect, no longer the missing curve of CTR01.

The sun on the rock behind the references, 14 October At 15:30 (middle) the rock around CTR01, CTR02 and CTR03 is bright. At 14:30 and 16:15 it is in shadow.

Mist, rain, and a shift that stays (mist and rain confirmed; cause of the shift unknown). In the dawn mist of the 14th, the targets jump by up to 17 mm. During the rain of the 23rd they jump by up to 18 to 19 mm. At 14:30 on the 23rd, the reference base changes: the targets keep a shift of about 1.5 to 2 mm (PR02, PR03 in the horizontal, PR06 in the vertical). It does not go away through the 24th.

Corrected curves in x of CTR02, CTR03, PR02, PR03 and PR12, 22 to 24 October After the rain, the curves sit at another level.

One-off jumps (cause unknown). PR02 and PR03 jump once on the 22nd at 08:45 (about 10 mm). All the targets jump once on the 24th at 14:15 and once at 15:15 (1 to 2 mm). Each time they come back at the next measurement.

Gaps. No measurement at all for 2.5 hours in the night of the 15th to the 16th, and for single measurements on the 16th and the 19th. The same gaps occur on the other stations.

What to do

  • PR09 is lost every night. Have its night visibility checked. It could get its own "problem target" label, like CTR01 and PR04.
  • Watch whether the shift after the rain of the 23rd comes back. If it doesn't, check CTR02 and CTR03 and the window in front of them.

Not established

  • The cause of the shift after the rain (water on a reference, or a mark on the window?).
  • Two short jumps of many targets lay inside removed labels and are not labelled: dusk on the 13th, 16:15-17:00, and PR07 on the 20th at 20:30. A new analysis of those hours would settle whether they deserve a label.

Requalification

Kept 31, rewritten 4, removed 9.

  • NORDEND-183, NORDEND-194, NORDEND-199, NORDEND-201, NORDEND-207, NORDEND-274 removed: night rejections of the left-edge targets, missing points only. The other targets meanwhile show only scattered single points of 5 to 10 noises, unrelated to them; the large jumps inside them carry their own labels.
  • NORDEND-188 removed: CTR01 rejected once; at most 5.4 noises on one point, nothing with a local level.
  • NORDEND-191 removed: CTR01 rejected once; nothing beyond 5 noises.
  • NORDEND-277 removed: cooling of the camera that the correction absorbs; nothing beyond 5 noises from its level.
  • NORDEND-186 rewritten: the sun on the rock disturbs reference CTR03, and PR13, PR12 and PR11 drop by 5 to 8 noises (rewritten on that effect).
  • NORDEND-208 rewritten: "the curves stay flat" was wrong. Five targets drop together at night by 5 to 7 noises (up to 11 against the night noise).
  • NORDEND-210 rewritten: same; six targets rise in the morning by about 6 noises and stay up.
  • NORDEND-214 rewritten: same; PR13 rises by 9 noises in the afternoon.

Labels

Start (UTC) End (UTC) Family Targets In a few words
13/10 07:45 14/10 05:00 Thermal displacement PR18, PR16, PR14 weak camera cycle, poorly absorbed
13/10 09:15 13/10 11:00 Visibility CTR04, PR16, PR15, PR06 reference poorly seen, targets jump
13/10 12:15 13/10 12:45 Visibility PR08, PR13, PR16 light dip, one point
14/10 05:00 14/10 06:15 Visibility all dawn mist, no valid point
14/10 06:15 15/10 07:15 Thermal displacement PR18, PR16, PR13 camera cycle
14/10 15:15 14/10 16:00 Visibility CTR03, PR13, PR12, PR11 sun behind the references
15/10 07:15 15/10 22:30 Thermal displacement PR18, PR16, PR06 camera cycle
15/10 22:30 16/10 01:15 No data all common gap on all stations
16/10 05:45 17/10 08:45 Thermal displacement PR18 camera cycle
16/10 13:15 16/10 13:45 No data all one missing measurement
16/10 14:00 16/10 14:30 No data all one missing measurement
16/10 19:00 16/10 23:00 Visibility CTR04, all targets reference weakens, targets jump together
17/10 08:45 18/10 07:15 Thermal displacement PR18, PR10, PR11 weak camera cycle
17/10 12:45 17/10 15:45 Visibility PR13, PR15, PR16, PR12 sun and clouds in the field
18/10 07:15 19/10 06:15 Thermal displacement PR18, PR15 weak camera cycle
19/10 06:15 20/10 07:00 Thermal displacement PR18, PR16 camera cycle
19/10 08:45 19/10 09:15 No data all one missing measurement
19/10 09:30 19/10 10:00 No data all one missing measurement
19/10 20:30 19/10 21:00 No data all one missing measurement
20/10 07:00 21/10 07:15 Thermal displacement PR18, PR16, PR14 camera cycle
20/10 15:15 20/10 17:15 Visibility PR07, PR09, PR15 faint target at dusk
20/10 23:45 21/10 01:45 Visibility PR14, PR15, PR16 three spots dim at night
21/10 07:15 22/10 06:30 Thermal displacement PR06, PR05, PR01 weak cycle, remainder at night
21/10 10:15 21/10 12:15 Visibility PR15, PR16, PR14 dim daylight on faint targets
21/10 14:15 21/10 17:15 Visibility PR15, PR16, PR14 faint targets at dusk
22/10 06:30 23/10 05:15 Thermal displacement PR01, PR12, PR05 weak cycle, camera does not come back
22/10 08:30 22/10 09:00 Displacement PR02, PR03 one jump, cause unknown
22/10 09:15 22/10 11:00 Visibility PR05, PR14 dim daylight on faint targets
22/10 12:00 22/10 14:00 Visibility PR07, PR16, PR08 dim daylight on faint targets
23/10 05:45 23/10 07:45 Visibility all dawn rain on the window
23/10 08:45 24/10 03:45 Visibility all rain day and night
23/10 14:15 23/10 14:45 Displacement CTR02, CTR03, all targets reference base shifts, no return
24/10 07:30 24/10 16:00 Thermal displacement PR13, PR06 weak cycle, small remainder
24/10 14:00 24/10 14:30 Displacement all whole image shifts once
24/10 15:00 24/10 15:30 Displacement PR08, PR15, PR16 base shifts once