A District-Level CAMS Forecast–ERA5 Reanalysis PM-Based Air-Quality Index Proxy and Meteorological Dataset for Khulna Division, Bangladesh (July 2015–December 2025)

Each district has 92,088 hourly timestamps. The dataset includes hourly PM2.5 and PM10 concentrations, rolling 24-hour PM-based index proxies, strict completeness companion indices, NowCast-based PM proxy indices, daily summaries, meteorological variables, quality flags, source provenance, and reproducibility materials. Known CAMS early-period lead-zero PM placeholder gaps were carefully handled.

Validated previous-day CAMS forecast step-24 records were integrated only when exact valid-time source data were available and spatial coverage requirements were satisfied. This process recovered 7,469 district-hour records per pollutant. After recovery, 721 PM2.5 and 721 PM10 concentration cells remain blank.

These remaining blanks represent unavailable source data and were not interpolated, estimated, or zero-filled. The air-quality values in this dataset should be interpreted as model-derived PM-based proxy indicators, not as official regulatory AQI, ground-monitor observations, or population-weighted exposure estimates. The dataset is intended to support transparent research on long-term particulate air-quality patterns, district-level meteorological conditions, environmental exposure screening, and regional air-quality studies in Khulna Division.

Provenance files, quality-control outputs, data dictionaries, boundary files, reproducible code, and checksum records are included to support verification and reuse.