Downwind distribution of an cancer relative risk about a point source: Single source with reactive and unreactive plumes

We examine the association between living from 1967 to 1969 within 5 km of steel furnace operations, petroleum refineries, and kraft and sulfite process pulp mills in Canada and the risk of all-cancer in 1993-1995. We find the risk of all-cancer is significantly associated with the early downwind distance. The adjusted R-2 of the correlation between risk and distance ranges upward from 0.75 to above 0.95. Relative risk increases to a maximum value, corresponding to direct exposure from emissions and then either declines slightly before rising again or continues to rise after a plateau. The RR never returns to 1. We find that additional risk is created within the plume. We find this additional risk in SIC sectors that have sulfur-rich emission plumes, a finding that implicates sulfur aerosols as a potential additional risk factor. Distance cannot be a surrogate for exposure when the source sector has SO2-rich plumes. We show that incremental RR can be extracted from cumulative RR to reveal additional sources of risk when residence distance from the sources is accounted for. This is the third in a continuing examination of retrospective exposure assessment (REA) with emission inventories.