Geographical, spatial, and temporal distributions of multiple indoor air pollutants in four Chinese provinces

Exposure to indoor air pollution from household energy use depends on fuel, stove, housing characteristics, and stove use behavior. We monitored three important indoor air pollutants-respirable particles (RPM), carbon monoxide (CO), and sulfur dioxide (SO2)-for a total of 457 household-days in four poor provinces in China (Gansu, 129 household-days; Guizhou, 127 household-days; Inner Mongolia, 65 household-days; and Shaanxi, 136 household-days), in two time intervals during the heating season to investigate spatial and temporal patterns of pollution. The two provinces where biomass is the primary fuel (Inner Mongolia and Gansu) had the highest RPM concentrations (719 microg/m3 in the single cooking/living/bedroom in Inner Mongolia in December and 351-661 microg/m3 in different rooms and months in Gansu); lower RPM concentration were observed in the primarily coal-burning provinces of Guizhou and Shaanxi (202-352 microg/m3 and 187-361 microg/m3 in different rooms and months in Guizhou and Shaanxi, respectively). Inner Mongolia and Gansu also had higher CO concentrations (7.4 ppm in the single cooking/living/bedroom in Inner Mongolia in December and 4.8-11.3 ppm in different rooms and months in Gansu). Among the two primarily coal-burning provinces, Guizhou had lower concentrations of CO than Shaanxi (1.2-1.8 ppm in Guizhou vs 2.0-13.3 ppm in different rooms and months in Shaanxi). In the two coal-burning provinces, SO2 concentrations were substantially higher in Shaanxi than in Guizhou. Relative concentrations in different rooms and provinces indicate that in the northern provinces heating is an important source of exposure to indoor pollutants from energy use. Day-to-day variability of concentrations within individual households, although substantial, was smaller than variation across households. The implications of the findings for designing environmental health interventions in each province are discussed.