LegacyAge depth estimates for the two Solsø cores

We joined the primary S5 and S6 tables from Li and colleagues' LegacyAge 1.0 PANGAEA deposit (2021, CC BY 4.0) to the named Neotoma sample depths at Solsø. The full source tables and their download receipts remain in the external bijux-pollenomics-data/legacyage/pangaea-933132 source store. The machine report records both source SHA-256 values, the exact Neotoma sample and age-claim hashes, source units and core identifiers.

Neotoma dataset	Core	Samples	S5 model depths	Exact sample-depth joins	Comparable calibrated ages	Refused calibrated ages
4445	SOLSOE81	66	492	66	65	1
4446	SOLSOE83	26	185	26	26	0

All 92 source samples have an exact S5 depth in centimetres; no rounding or interpolation was used. The sample ledger retains each Neotoma calibrated point age and the corresponding S5 minimum, maximum, median, mean and reported uncertainty fields. The 1981 sample 352996 at 98 cm has a calibrated Neotoma point age of 7 BP with a negative younger bound. Its Neotoma claim remains refused (negative_bp) in the refusal ledger, despite having an exact S5 depth and a LegacyAge interval. This is the same source-age issue recorded in the published EPD age correspondence.

For SOLSOE81, 57 of 66 Neotoma point ages lie within their matched S5 95% ranges and nine lie outside. The 57 include the refused negative-bound Neotoma claim, leaving 56 of 65 comparable claims inside. The LegacyAge mean minus Neotoma point age spans −635 to +378 years (median +16). For SOLSOE83, all 26 point ages lie within their matched S5 ranges; the mean difference spans −471 to +158 years (median −3.5). S6 separately describes the 1981 original chronology as partially or completely outside the LegacyAge 95% confidence intervals and the 1983 original chronology as within them. These are different age models; row-wise interval membership is a descriptive source comparison, not an inference about a joint age posterior.

This join does not identify which samples made up LandClim II's published 79-sample Solsø sequence, prove its historical age-model selection, provide sample-age basin geometry, or admit LRA input. It can support a separately declared observed-pollen chronology sensitivity with its own fixed prediction and matched-support accounting.

The producer checks source hashes and core/sample ownership, then writes the sample ledger, refusal ledger and report. Replay from the repository root into an absent disposable output directory:

PYTHONDONTWRITEBYTECODE=1 .venv/bin/python \
  docs/report/countries/denmark/solsoe-source-lineage/legacyage-chronology/replay.py \
  --repository . \
  --source-dir ../bijux-pollenomics-data/legacyage/pangaea-933132 \
  --output artifacts/jobs/solsoe-legacyage-depth-join-check

The two primary sources are the S5 chronology table (66,989,363 bytes; SHA-256 bbe0e8d5854acf091a11737ddedc10d5656f8a25c8eeb9637b85f9e69a37cb41) and S6 chronology comparison (438,320 bytes; SHA-256 2a058f08472199b0545e4562e6860371dfaba7ce74140b1e0b4dbdce0ce55300). Cite Li, Chenzhi; Postl, Alexander; Böhmer, Thomas; Dolman, Andrew M; Herzschuh, Ulrike (2021): Harmonized chronologies of a global late Quaternary pollen dataset (LegacyAge 1.0), PANGAEA, doi:10.1594/PANGAEA.933132.