Stochastic simulation of diurnal variations of CO, NO and NO2 concentrations in occupied residences

Abstract

A stochastic approach to the problem of incorporating variable emission events in indoor air quality models is proposed. A nonstationary Poisson process is used to account for the occurrence of range-top burner use. The combination of this emission event sequence with a simple one-compartment mass-balance model results in output that qualitatively agrees with measured concentrations in two occupied townhouses. Improved monitoring of stove usage times, gas flow rates and emission factors under field conditions would allow more effective estimation of model input parameters and more accurate prediction of concentration distributions.