An investigation of the parameters influencing the determination of the number of particulate matter sources and their contribution to the air quality of an indoor residential environment

Indoor air quality depends on the presence of both indoor and outdoor particle sources each of which produces different particles' size distribution that may have mortality and morbidity effects. Positive Matrix Factorization (PMF) is a mathematical (statistical) procedure for identifying and quantifying the sources of air pollutants at a receptor location. A critical step in PMF is the number of factors determination and the present study aims at discussing this critical issue, by applying PMF on particles size distribution measurements data in a residential environment, in Athens, Greece. A main focal point of the present research is the investigation of the temporal behaviour of the particles size, as recorded in the time series, closely relating the averaging period of the utilised data with the number and type of factors in the PMF. The analysis is based on the estimation of the spectral properties of data and estimation of the integral time scale using the autocorrelation properties of the series. Furthermore, different factor analysis techniques have been applied, namely the rotated Principle Component Analysis (rPCA) and the Independent Component Analysis (ICA) and the results have been compared to PMF results.