Predicting black smoke levels from deposit gauge and SO2 data to estimate long-term exposure in the United Kingdom, 1956–1961

Abstract Background In the UK air quality has been monitored systematically since 1914, providing valuable data for studies of the long-term trends in air pollution and potentially for studies of health effects of air pollutants. There are, however, challenges in interpreting these data due to changes over time in the number and location of monitored sites, and in monitoring techniques. Particulate matter was measured as deposited matter (DM) using deposit gauge monitors until the 1950s when black smoke (BS) filters were introduced. Estimating long-term exposure to particulates using data from both deposit gauge and BS monitors requires an understanding of the relationships between DM, SO 2 and BS. Aims To explore whether DM and/or SO 2 , along with seasonal and location specific variables can be used to predict BS levels. Methods Air quality data were abstracted from hard copies of the monthly Atmospheric Pollution Bulletins for the period April 1956–March 1961 for any sites with co-located DM, SO 2 and BS data for three or more consecutive years. The relationships between DM, SO 2 , and BS were assessed using mixed models. Results There were 34 eligible sites giving 1521 triplets of data. There was a consistent correlation between SO 2 and BS at all sites, but the association between DM and BS was less clear and varied by location. Mixed modelling allowing for repeat measurements at each site revealed that SO 2 , year, rainfall and season of measurement explained 72% of the variability in BS levels. Conclusions SO 2 can be used as a surrogate measure for BS in all monitoring locations. This surrogate can be improved upon by consideration of site specific characteristics, seasonal effects, rainfall and year of measurement. These findings will help in estimating historic, long-term exposure to particulates where BS or other measures are not available.