Incorporation of Radon-222 as a parameter in ME-2 to improve apportionment of PM2.5 sources in the Sydney region

•Ion Beam methods used to characterise fine particulate dataset.•ME-2 technique used to combine wind direction, wind speed, Radon-222 and PM2.5.•The Radon-222 parameterisation was compared against the wind speed parameterisation.•Radon-222 parameterisation produced superior results for large number of cases.

In recent years source apportionment of observed PM2.5 has been improved by incorporating meteorological information as additional factors in receptor modelling studies using ME-2. In this study we replace one of these meteorological factors, namely, parameterisation by wind speed, with a parameterisation based on hourly observations of the naturally occurring terrestrial gas Radon-222 (radon), and compare results of the two parameterisations over five years at an inland site in the Greater Sydney Region.The efficacy of the wind speed and radon parameterisation techniques is assessed by comparing regressions between the daily contributions from identified elemental fingerprints estimated from the wind speed and radon multi-linear models against those obtained from the corresponding bi-linear model (while the two models are solved simultaneously).The radon parameterisation yielded improved regressions for all source fingerprints, most notably Smoke and Autos (r2 = 0.67 and 0.65, respectively, compared to 0.57 and 0.47 when the wind speed parameterisation was used). Both parameterisation schemes were equally effective in attributing PM2.5 to wind sectors known to contain sources characterised by the observed fingerprints.Our findings demonstrate that incorporating radon as a parameter in ME-2 can lead to an improved PM2.5 source apportionment than that obtained using meteorological parameters alone, particularly for inland sites with distributed sources.