Quality control
COFECHA-style checks at the current placement
vs
Placement
1135–1276
uncommitted
Series r
-0.018
99% crit 0.195 · n 142
Segments flagged
4/4
50-yr lagged 25
Ring flags
0
missing / false
Outliers
0
> +3 / < −4.5 SD
Skeleton agreement
35%
22/62 narrow rings
AR order
0 / 3
series / master
Skeleton plot
narrow rings · specimen ↑ · master ↓ · scores ≥ 3
Segment correlations
A = r below 99% critical · B = higher r at another shift (±10 yr)
| Segment | r at dated position | r | crit | best shift | flag |
|---|---|---|---|---|---|
| 1150–1199 | -0.08 | 0.33 | +9 (r 0.20) | AB | |
| 1175–1224 | -0.14 | 0.33 | +6 (r 0.45) | AB | |
| 1200–1249 | -0.10 | 0.33 | +5 (r 0.17) | AB | |
| 1225–1274 | 0.05 | 0.33 | +2 (r 0.12) | AB |
Ring-level flags
No missing, false or outlier rings detected at this placement.
Workspace screening (4)
| Specimen | r | Seg A/B | Ring flags |
|---|---|---|---|
| SG-OAK-014 | -0.018 | 4/4 | 0 |
| ALP-PCAB-003 | -0.057 | 3/3 | 0 |
| ARCH-QU-W07 | 0.067 | 2/2 | 0 |
| PANEL-BQ-02 | 0.007 | 6/6 | 0 |
Each specimen at its current placement against its assigned master.
Method
After COFECHA (Holmes 1983; Grissino-Mayer 2001). Series and master are log-transformed, high-pass filtered with a Gaussian of 50% response at 32 yr (an approximation of COFECHA's 32-yr spline), and prewhitened with an AR model (order ≤ 3 by AIC).
Missing/false rings: whole-series breakpoint search for a ±1-yr shift, accepted when the tail correlation passes the 99% critical r and improves by > 0.2.
Skeleton plot after Stokes & Smiley (1968): ring width relative to the ±5-ring local mean, scored 0–10.