Skip to content

Amsoldingersee publication-input source review

LandClim II metadata row 1059 identifies Amsoldingersee, source code E#69, 23 pollen samples, and an author chronology. The row lies outside the 17 author-prepared examples. Its NR restriction label is a workbook value, not a per-dataset redistribution decision.

The source-bound report replays a 21 September 2026 Neotoma API capture from private source custody. Site 2973 has three current pollen datasets: 3873 (AMSOPO1, 23 sample IDs), 3874 (AMSOPO2, 24), and 3875 (AMSOPO3, 52). Dataset 3873 matches the workbook sample count but its chronology response has zero records. The other two return radiocarbon-year age information. None of the three dataset, download or chronology responses contains E#69. A site-name, coordinate or sample-count match cannot identify the paper's historical core, count revision or age model.

One selected publication row remains unrecovered; zero of three candidate cores is reproduction-ready. The missing inputs are an author or ALPADABA E#69-to-core mapping, the exact historical 23-sample count revision, the author age-depth output, and the taxon-to-PFT transformation. No R input, vegetation accuracy or source-use permission is inferred from this review. The full 1,128-row inventory retains the broader denominator; this review resolves the refusal evidence for one row without changing its admission status.

The source-input manifest binds the 11 consumed raw API responses, their two acquisition receipts and the selected publication row. Historical run plans, logs and a superseded public-note check are not inputs to this source decision.

To replay into an absent artifact directory with the private source store:

PYTHONDONTWRITEBYTECODE=1 .venv/bin/python -m \
  bijux_pollenomics_dev.source_completion.landclim_amsoldingersee_recovery \
  --repository . \
  --source-root /path/to/private/landclim/amsoldingersee-source-review \
  --output artifacts/jobs/landclim-amsoldingersee-source-review/replay

The producer checks all 14 consumed private files against the source manifest, all 11 HTTP response hashes and status records, the exact author example metadata, and the observed candidate fields before writing the report. The raw API responses remain outside the repository.