Runs 7–8 · Investigation 2Run 7 finished 04:21 · Run 8 finished 04:360.9 CU of 0.9 reserved · meridian-hpc
Run 7 forecast at A3, observed · Run 8: post-event capacity , factor of safety
Run 8 fits the tilt and the tension peak with clay about 6 % weaker near the caisson tip (su −6 %). That clay update is recorded, not adopted: monitoring keeps twin v2 until gate G3 accepts it.
EElidyneexplanation · answered from records04:38grounded in 5 records
Did the anchor hold through the shaking?
At 04:13:40, 79 s after the seabed trigger, the anchor + mooring twin v2 forecast a top-tension peak of from the seabed motion recorded so far. The load cell read at 04:14:02: a residual of , inside the twin's 95 % band. Caisson tilt went from to , half the 1.0° advisory band, and has not moved since 04:15.
Run 8 re-ran the twin with the full recorded seabed motion. Post-event capacity is against the design value: a factor of safety of against the 185 kN cyclone demand, where the owner's criterion is 2.0. The fit asks for clay about 6 % weaker near the tip, which is what shaking at 212 km should do to soft pelagic clay of sensitivity 3.
The anchor held, inside the twin's bandThe clay update (su −6 % near the tip) waits for gate G3 with A. Rao.
Figure 1. Mooring top tension, 04:10–04:20, against the twin v2 forecast and its 95 % band. Shaded column: shaking at A3, 04:12:21–04:13:50. The dashed band is Run 8, the twin re-run with the recorded seabed motion, added at 04:36. The 100-year cyclone design peak is 185 kN.
Hover the figure, or focus it and use the arrow keys, to read one sample.
Checked and set aside
Load-cell fault: the redundant cell LC-2 read 94 kN at 04:14:02, within 2 % of LC-1.
Buoy excursion: GNSS offset 38 m at 04:14, inside the 120 m watch circle; the line did not go slack.
Tilt sensor: both inclinometer axes agree within 0.05°, steady at 0.5° since 04:15.
Not ruled out: further softening near the tip from aftershocks. The largest so far is Mw 5.4 at 04:47, 0.04 g at the seabed; Anchor integrity watch stays on the tilt and tension bands.
Updated 04:36 by Run 8 of Investigation 2: post-event capacity 596 kN, factor of safety 3.2. The clay update stays proposed until gate G3.
One press approves the whole plan. Nothing runs until you execute it.
EElidynemonitor review04:46
Shadow · never acts
Tsunami watch v2 (proposed, running in shadow since 06:00 on 15 Sep · never acts)
Compared with v1, live, on the same 5 streams over the last 24 h (04:52 on 21 Sep to 04:52 today).
Last 24 h
v1 live
v2 shadow
Alerts raised
5
2
Later cleared as noise
3
0
Valid alerts missed
0
0
Tsunami arrival raised
04:32:10
04:31:55 · 15 s earlier
valid alertcleared as noise
›What changed in v2 · 3 rule changes
Detection: 2 samples above 3 cm with a rise of more than 1 cm per 15 s, instead of 3 samples above 3 cm. Today's arrival would have been raised 15 s sooner.
Seabed trigger: two axes above 0.15 g, or one axis for 3 consecutive samples. Single-axis spikes stay quiet.
De-tiding: tide model refitted on 180 days of BPR-3 records, so spring-tide residuals stay under 3 cm.
›All 5 alerts, side by side
When
What
v1
v2
Outcome
Promotion needs your approval · the L2 grant carries over unchanged
Promote Tsunami watch v2 to live? v1 stops raising alerts and moves to shadow for 7 days. v2 keeps the same standing approval: L2 for BPR-3 event mode only, L0 for everything else, may read 5 streams, ≤ 0.2 CU per firing.