The origin and day-of-week dependence of photochemical ozone episodes in the UK

Analyses are presented which focus specifically on photochemical ozone events when hourly mean levels of ozone at UK monitoring sites have reached or exceeded the information threshold of 90 ppbv. The geographical origins of such episodes are determined from an analysis of the trajectories associated with elevated ozone levels at 14 rural sites and six urban sites in the UK over the period 1992-1999. The analysis shows that the highest levels of ozone occur under summertime anticyclonic conditions, when the back trajectories have 'looped' over mainland Europe and arrive in the UK from a broadly easterly or southeasterly direction. Monitoring data from the same sites over the period 1989-1999 clearly demonstrate that ozone episodes are more prevalent at the end of the week, with the greatest numbers of hours greater than or equal to 90 ppbv occurring on Fridays. Relative to Fridays, the number of hours exceedence on the other days are: Sundays 51%, Mondays 50%, Tuesdays 54%, Wednesdays 39%, Thursdays 70% and Saturdays 82%. The observed day-of-week dependence is believed to result from the temporal dependence in the emissions of the precursor VOC and NOx (which are greater on weekdays) and the multi-day timescale required for chemical processing and transport to lead to elevated ozone levels under photochemical episode conditions, To test this hypothesis, a photochemical trajectory model, incorporating temporal variations in precursor emissions, has been used to investigate the formation of ozone along 96 h trajectories arriving at eight southern UK rural sites, for the meteorological conditions of an ozone episode on 31 July 1999, As a direct result of imposing a day-of-week variation in emissions, the model is able to recreate some of the features of the observed general pattern of 90 ppbv exceedences with day of week. In particular, Friday is calculated to be the day with the highest peak ozone mixing ratio at 7 of the 8 sites. The calculated peak mixing ratios on each day at each site are used to infer the number of hours exceedence of the 90 ppbv threshold, using a relationship derived from archived monitoring data. The inferred hours exceedence of the 90 ppbv threshold at all sites combined is calculated to be a factor of two greater on Friday than it is on Monday or Tuesday. (C) 2002 Elsevier Science Ltd. All rights reserved.