Data and code for Scoring populations and model-ranking sensitivity in sugar beet harvest forecasting: an auditable case study

Version 13.0 of the data and code package supporting the manuscript “Scoring populations and model-ranking sensitivity in sugar beet harvest forecasting: an auditable case study” (submission identifier RSOS-262316). The package contains author-prepared analysis inputs derived from the cited source datasets, saved prediction tables and numerical outputs used in the reported analyses, model settings, analysis and figure-generation scripts, integrity checks and checksums, and a numerical transcription of the Montana source tables with provenance. The third-party German multi-environment raw files and the Braunschweig FACE source workbook are not redistributed in this package. The German source is available through https://doi.org/10.18174/odjar.v11i0.18784. The FACE data are available from Harvard Dataverse at https://doi.org/10.7910/DVN/PNXH43 and are described in https://doi.org/10.18174/odjar.v5i0.16225. Consult the source records and the package's source-data instructions for access, attribution and original terms. The package's reconstruction instructions identify which steps require these external files. Synthetic outputs are explicitly labelled and are not appended to observed crop data. The Montana transcription was checked against source-page images; it was not independently transcribed a second time by another human. This record preserves materials for reproducing and auditing the reported analyses; it does not establish prospective or operational forecasting performance. Author-written code is licensed under the MIT License. Author-created analysis tables, numerical outputs, figures and documentation are licensed under Creative Commons Attribution 4.0 International (CC BY 4.0). These grants apply to the authors' own contributions. Original third-party materials retain their original terms and attribution requirements; they are not relicensed by this record. See the accompanying licence and source-attribution files for the scope of each licence.