Volcanoes/aerosol: When the analysis includes both the El Chichón and Pinatubo eruption periods (i.e. starting in 1979), you should not use a single aerosol proxy. El Chichón erupted during a period of relatively low chlorine content, while Pinatubo erupted during a time of significantly higher chlorine content. Tie and Brasseur [1996] showed that the impact of an eruption is sensitive to the chlorine content. The initial reaction is N₂O₅ + H₂O →2HNO₃ on the surfaces of the sulfate particles from the eruption. The result is a reduction in NO_x that leads to an increase in ozone. An important secondary effect is that the reduction in NO_x leads to a decrease in the conversion of ClO_x to chlorine nitrate, ClONO₂, thus increasing the effectiveness of chlorine catalytic loss. Pinatubo erupted when chlorine content was high and the chlorine effect dominates the NO_x effect leading to a clear decrease in ozone. Tie and Brasseur [1996] estimated that El Chichón erupted when the chlorine content was such that the two effects nearly cancelled and there should be little net effect on ozone. Thus, statistical fitting should have independent terms for these two volcanoes. Using only one term will probably lead to getting half of the correct answer for Pinatubo and a false answer for El Chichón.