Observer redundancy and biodiversity visibility: analysis code and summary results

This record contains the analysis code and summary result tables for the study “Observer redundancy shapes the resilience of global biodiversity knowledge”.

The archive provides ordered Python and R scripts used to process GBIF-mediated biodiversity occurrence records, construct observer- and region-year analytical panels, quantify observer redundancy and concentration, fit predictive models of next-year biodiversity visibility loss, evaluate model generality across monitoring regimes, and simulate targeted observer-support scenarios. The accompanying CSV files contain the summary outputs used for manuscript reporting, including model validation metrics, transfer validation results, redundancy threshold summaries, structured-null concentration summaries, and targeted-support scenario results.

Raw GBIF-mediated occurrence data are not redistributed in this archive. Users should obtain the original occurrence data through GBIF and cite the corresponding GBIF download DOI(s). The files deposited here are intended to support computational transparency, reproducibility of reported summary results, and reuse of the analysis workflow.

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

Parquet analysis tables are included, these are aggregated or processed analytical tables generated after quality control, spatial aggregation, and feature construction. They are provided to facilitate reproducibility of the modelling workflow without redistributing the full raw GBIF occurrence download.