Table 2. Dendro-14C-wiggle-match placements for the Anatolian tree-ring series.
| Dating Model | Elements /dates | ring 776 Date BCE | 68.2% hpd range | 95.4% hpd range | Amodel | Aoverall | Outliers | |
|---|---|---|---|---|---|---|---|---|
| Gordion Scheme Relative Years (RY) | ||||||||
| μ±σ | M+ | Date BCE | Date BCE | |||||
| 1. GOR RY776.5–1764 | ||||||||
| 128/128 | 1738±1 | 1738 | 1740–1737 | 1741–1736 | 0.8 | 0.8 | 25 | |
| 2. = 1 minus outliers | ||||||||
| 103/103 | 1738±1 | 1738 | 1739–1736 | 1741–1735 | 123.4 | 118.4 | 1 | |
| 3. = 2 minus outlier | ||||||||
| 102/102 | 1738±1 | 1738 | 1740–1737 | 1741–1735 | 141.4 | 133.9 | 0 | |
| 3a. = 3 with ΔR 0, 10 (μ±σ = 6.9±2.3) | ||||||||
| 1738±2 | 1738 | 1739–1736 | 1741–1734 | 247.7 | 233.2 | 0 | ||
| 4. GOR RY776.5–1145.5 | ||||||||
| 50/50 | 1732±3 | 1732 | 1736–1730 | 1739–1729 | 10.9 | 12.4 | 7 | |
| 5. = 4 minus outliers | ||||||||
| 43/43 | 1734±2 | 1734 | 1737–1732 | 1737–1726 | 97.7 | 98.8 | 0 | |
| 6. POR | ||||||||
| 18/51 | 1753±7# | 1754 | 1759–1751 | 1762–1740 (91.4%), 1736–1730 (4%) | 92.6 | 103.1 | 0 | |
| 6a. = 6 with ΔR 0, 10 (μ±σ = -5.6±6.4) | ||||||||
| 1754±8 | 1756 | 1760–1753 | 1764–1746 (89.6%), 1735–1725 (5.8%) | 118.1 | 125 | 0 | ||
| 6b. POR (OxA & VERA separate) | ||||||||
| 1753±6 | 1755 | 1759–1752 | 1762–1740 (93.2%), 1736–1732 (2.2%) | 100 | 107.3 | 0 | ||
| 7a. KUL-KBK-ACM = MBA (KUL and ACM samples separate) | ||||||||
| 39/103 | 1746±8 | 1744 | 1749–1738 | 1771–1734 | 4.3 | 13.1 | 6 | |
| 7b. 7a but with KUL+ACM combined, minus 1 sample* | ||||||||
| 34/102 | 1746±10 | 1743 | 1749–1737 | 1773–1764 (9.8%), 1757–1773 (85.6%) | 5.5 | 18.8 | 5 | |
| 8a. = 7a minus outliers | ||||||||
| 33/90 | 1744±4 | 1744 | 1748–1739 | 1752–1736 | 137.2 | 142.1 | 0 | |
| 8b. = 7b minus outliers | ||||||||
| 29/90 | 1745±4 | 1745 | 1749–1741 | 1753–1738 | 121.3 | 128.1 | 0 | |
| 8c. = 8a with ΔR 0, 10 (μ±σ = 3.4±5.7) | ||||||||
| 1742±5 | 1741 | 1746–1736 | 1752–1732 | 170.7 | 167.9 | 0 | ||
Calendar date (BCE) placements for the dendro-14C-wiggle-matches of the Gordion, Porsuk and MBA tree-ring chronologies (Fig 3) employing IntCal13 and OxCal [43, 61, 62] expressed in terms of RY776 following the previous chronology based on the Gordion RY sequence and dates published [21–23, 31, 40, 41]. The models consider first all data in each series, and then excluding outliers if present (RScaled outlier model in OxCal [63]) and considering possible 14C offsets (ΔR [62]). Typical outputs are shown (see also Figures J-N in S1 File). The OxCal Amodel and Aoverall values should be about ≥60 to indicate satisfactory agreement within the model. Elements are the separate dendro-sequenced analytical units in the models–these may comprise one or more dates; where plural these dates on the identical tree-rings are employed as a weighted average [60]. Note: RY776 is simply an arbitrary choice (but used since it was the reference point for previous Gordion-based dates) to allow comparison of the different chronologies. If the previously published crossdates and positions were correct (e.g. [21–23, 31, 40, 41]), then all the chronologies should yield approximately the same calendar age ranges for RY776. They do not. The tree-ring series do not all include an RY776 –this year is extrapolated in such cases to enable the comparison in terms of a single year.
+ M = median value
# The probability distribution for the Porsuk series is not approximately symmetrical–contrast all other cases–the mode of the distribution is 2 years older than the mean, and 1 year older than the median–at 1755 BCE for Model 6. We use 1755 BCE as the best fit point.
* OxA-30907 for RY607-615 had divergent δ13C values recorded by the AMS versus those independently measured on an MS, and so is considered inherently less than entirely reliable and so was excluded in run 7b (but included in run 7a).