Emission and activity uncertainty limits source apportionment for PM2.5 mass in residences

Deterministic application of computational simulations of PM2.5 mass concentration in residences, informed by occupant diaries, allows evaluation of the uncertainty from emission rates and activity character. Uncertainty is generated by emission rates through varying methods to quantify rates of a given activity, whilst uncertainty is generated by activity character by the fact that participant surveys typically only ask for what activity is occurring, thereby omitting information on how it is occurring. Using the PyCHAM (CHemistry with Aerosol Microphysics in Python) box model alongside measurements for contrasting case studies from inhabited family homes in Bradford (UK), we show that these uncertainties are sufficiently great to undermine apportionment of key sources of PM2.5 in residences, motivating future work to reduce uncertainty.