An open environmental-justice method for prioritising protection of schoolchildren from air pollution in Central Asia — data and code

Compact derived datasets and the complete analysis pipeline for an open, transferable environmental-justice method that prioritises school-level air-pollution protection in data-scarce cities, demonstrated across six Central Asian capitals (Tashkent, Almaty, Astana, Bishkek, Dushanbe, Ashgabat) and transferred out-of-region (Accra, Kathmandu, Lima).Contents (see ZENODO_MANIFEST.md and CLAIM_INDEX.md): per-school exposure surfaces, building-age, wealth (Meta RWI), under-20 child-density and NO2 layers; the injustice/retrofit-priority index and its robustness runs (random weightings, VIIRS and GRDI deprivation cross-checks, surface-parameter sweep, indoor/outdoor invariance); measured on-school validation; reference and municipal-network PM2.5; health (GBD 2021) inputs; global-applicability and OpenAQ reference-coverage tables; and all Python code. Raw third-party rasters (ACAG, GRDI, WSF, WorldPop) are NOT redistributed; each script fetches them from their public source, so the pipeline is fully reproducible from open data.Data CC-BY-4.0; code MIT. Associated manuscript submitted to npj Urban Sustainability.v2.1.0 (data-integrity update). Five derived files under outputs/temporal/ in the previous version were stale artefacts computed from the superseded low-cost optical sensor (OpenAQ station 4902926) rather than the reference monitor: analysis_summary.csv, school_exposure_detailed.csv, detailed_temporal_analysis.csv, policy_recommendations.csv and pm25_daily_means.csv. Root cause was a path bug that left their generator scripts pointing at a directory that no longer existed after a repository reorganisation, so they were never regenerated. The generators are fixed and the four reproducible files are regenerated from the reference-grade U.S. Embassy record (station 8881; n=8,301; mean 37.8 ug/m3); pm25_daily_means.csv is withdrawn as redundant with outputs/reference/us_embassy_2022_2023_daily.csv. No manuscript claim was affected -- none of these files feeds any reported number (see CLAIM_INDEX.md). Also adds the named Tashkent school retrofit-priority list (data/pipeline/indices/tashkent_school_priority_named.csv) and an updated DATA_CODEBOOK.md.