Observer redundancy and biodiversity visibility: analysis code and summary results

This record contains the custom analysis scripts and summary result tables associated with the study “Observer redundancy shapes the resilience of global biodiversity knowledge”.

The archive includes ordered Python and R scripts used for data ingestion, quality control, terrestrial and inland filtering, observer-panel construction, region-year and species-cell-year analytical panel construction, structured null-model fitting, predictive fragility modelling, model validation, transfer validation, targeted-support simulation, and figure/table generation.

The accompanying CSV files provide summary outputs used for manuscript reporting, including validation metrics, regime-specific model performance, transfer-validation results, redundancy-threshold summaries, threshold-stability summaries, observer-concentration summaries, and targeted-support scenario results.

The raw occurrence data analysed in the study were obtained from the public iNaturalist open data snapshot. Raw iNaturalist occurrence records and large intermediate Parquet analysis tables are not redistributed in this Zenodo record. The deposited files are intended to support transparency, reproducibility of reported summary results, and reuse of the analysis workflow.

Folder structure:- scripts/: ordered analysis scripts- results_csv/: summary CSV result tables used in manuscript reporting