High-resolution modelling of residential heating emissions in Poland

Residential heating remains a dominant source of air pollutant emissions in Europe, especially in countries like Poland, where solid fuels such as coal and wood are widely used for household heating. Emissions from this sector significantly contribute to elevated particulate matter concentrations (PM2.5, PM10), benzo(a)pyrene (BaP), and other harmful pollutants, often exceeding regulatory air quality standards and posing serious public health risks. Accurate, spatially resolved emission inventories are crucial for effective air quality management and policy development, yet uncertainties in fuel type allocation and incomplete coverage of emission sources have limited traditional approaches. This study presents an enhanced bottom-up inventory for residential heating emissions in Poland, introducing two major methodological advancements: the integration of the Central Register of Building Emissions (CEEB)-a national database requiring annual declarations of primary heating fuel types by building owners-and the inclusion of emissions from domestic hot water production. The baseline methodology combined country-wide and local datasets, including building characteristics (from BDOT10k), heating degree days (GEM-AQ), fuel mix (municipal statistics), gas usage, building age, and heat distribution network data. The updated approach geocoded 12 mln CEEB declarations to assign precise heating fuel types and system characteristics to individual buildings, distinguishing between heating and hot water systems. Fuel shares for zero-emission sources, gas, oil, coal, and wood were determined at high spatial resolution. Hot water emissions were calculated using standardized energy demand per household, adjusted for building type and size. Two emission scenarios were evaluated in the GEM-AQ air quality model at 2.5 km resolution: one using the base approach, and one incorporating CEEB data and hot water emissions. Model validation against reference station measurements demonstrated that the enhanced inventory notably improved model performance, reflecting more accurate spatial distribution and quantification of residential emissions. These advancements provide a robust foundation for air quality modelling.