Estimation of indoor radon concentration and automatic risk classification in multi-storey buildings

Radon ( 222 Rn) is a carcinogenic noble gas that accumulates indoors, posing significant public health risks, particularly lung cancer. This study presents a novel conceptual model and algorithm for estimating indoor radon concentration (IRC) in multi-storey buildings by integrating three primary sources: underlying lithology (Ql), active faults (Qf), and building materials (Qb). A porosity-based weighting system (n = 0–1) was developed to account for radon transport through different building elements (floors, walls, ceilings). The algorithm calculates radon concentration at ground floor (Q 1 = Ql + Qf + Qb) and first floor (Q 2 = Qb), with porosity parameters controlling gas migration. Using a comprehensive dataset of 117 buildings from İzmir, Türkiye, the model was validated. Statistical analysis showed an overall mean winter-period IRC of 210.2 Bq/m 3 , with the highest district-level mean observed in Buca (236.2 Bq/m 3 ), followed by Bornova (220.9 Bq/m 3 ). Approximately 15.3% of the surveyed buildings exceeded the EU-recommended reference level of 300 Bq/m 3 during the winter measurement period. The proposed risk classification scheme categorized 15.3% of the surveyed buildings as high-risk during the winter measurement period. The proposed physics-informed framework offers a cost-effective, interpretable tool for proactive radon risk assessment and mitigation planning.