A harmonized 30-year hourly air-quality dataset for Tehran (1995–2024): multi-agency integration of ground monitoring with satellite-based validation

Long-term, quality-controlled ground measurements remain the reference standard for air-quality assessment. They are also scarce in many of the regions that carry the heaviest pollution burden, where the records that do exist are often split across institutions, hard to access, or in danger of being lost. Tehran is a case in point. The city, one of the most polluted megacities in the Middle East, is monitored by two institutionally separate authorities: the Tehran Air Quality Control Company (AQCC) and the national Department of Environment (DOE) with its provincial directorate. Their records have been archived and distributed independently of one another, and only the recent AQCC data are public. We present a harmonized, quality-controlled archive of predominantly hourly observations for Tehran covering 1995 to 2024. It brings together five provenance streams: the recent and historical records of both authorities, plus measurements for 1995–1996 that were hand-digitized from a historical technical report. The combined record holds more than 17.6 million quality-controlled observations of six pollutants (CO, NO₂, O₃, SO₂, PM₁₀, and PM₂.₅) at 49 monitoring stations. CO is reported in parts per million by volume, the other gaseous species in parts per billion by volume, and the particulate fractions in micrograms per cubic metre. Every record passed through one scripted, fully auditable pipeline. That pipeline harmonized the heterogeneous source formats, reporting units, and timestamp conventions, settled documented data-quality issues, and stripped out no-data codes, physically implausible values, and erroneous duplicate segments. We applied no statistical outlier removal, no gap-filling, and no smoothing, so the released record stays close to what was observed. For each observation we keep the provenance, recording resolution, and operating period as explicit attributes. The record ships in both long and wide tabular forms, with a station inventory and the full processing logs. We validated it against satellite observations. Per-station rank correlations between the surface measurements and co-located TROPOMI tropospheric columns of NO₂, SO₂, and CO are positive over the 2018–2024 overlap. The particulate-matter and total-column-ozone comparisons are reported in full but, for physical reasons we document, are not treated as validation of the surface record. To our knowledge, this is the first openly deposited, multi-pollutant, multi-agency, provenance-preserving hourly air-quality archive for Tehran spanning 1995–2024, and it is intended as a reusable resource for epidemiological, atmospheric, exposure-assessment, and policy research in a high-exposure, data-scarce region.