Atmospheric measurements of peroxyacetyl nitrate and other organic nitrates at high latitudes: Possible sources and sinks

,

A small correction (3 to 7%) was applied to account for trace quantities of methyl nitrate that were nearly always present.A two-step dynamic dilution system was used to generate low-ppt mixing ratios of PAN.A measurement sensitivity of better than 5 ppt PAN was achievable with a 100-mL air sample.Insmnnent precision and accuracy are estimated to be +10% and +25%, respectively.Additional details of this and similar techniques can be found elsewhere [Singh andSatas, 1983, Joos et al., 1986;Fehsenfeld et al., 1987;Singh et al., 1990a].Methyl nitrate mixing ratios were calculated by using gas phase cou!ometery [Lillian and Singh, 1974].Preliminary tests suggest that this should be accurate to +50%.Although no measurements from the stratosphere have been' published to date, these data suggest that substantial concentrations of PAN in the upper troposphere and lower stratosphere may exist.Ocean-land variabilities.It was repeatedly observed that in the boundary layer, PAN over land was higher than over water.Figure 14  NOy did not perceptibly change during this transition.It is also noted that in virtually all such instances the land air temperature was warmer than the marine air; hence PAN was actually less stable over land than over water.A more plausible explanation lies in the more rapid vertical mixing, caused by warmer land temperatures, with the reservoir aloft.This is further supported by the observation that 0 3 was also about 2 ppb higher over land than over water.
Forest fire plumes.Figure 15 shows PAN enhancement at around 2 km inside a haze layer intersected during the overland mission 14 (July 26/27).Inside the haze layer NO, 03, and CO showed concurrent increases.Other trace species such as C2H 2 and C2H 6 also showed significant excursions, while synthetic species such as C2C14 were unchanged.It is likely that this haze layer consisted of residual smoke from forest fires that had been burning in the area during the previous several days [Drewery, 1988].Within the plume a significant fraction of

Flight
ID) was f'dled and held at ice-water temperature in a specially constructed dewar.Air flowed over this tube at a known rate.The PAN in the exit stream was measured by using an NOy detector.A nylon filter was inserted in line to remove any traces of nitric acid.
shows one such example (mission 19, August 2/3) when the aircraft flew over water and land at about 200-m altitude.PAN levels nearly doubled when the land mass approached.A similar increase when PAN levels increased from about 12 ppt to 26 ppt was also observed during mission 17.The reasons for this rapid transition are not likely to be associated with the additional soil sources of NOx, which are very small [BaSin et al., this issue] and indeed both NO x and

Fig. 14 .
Fig. 14.PAN at the sea-land interface in the sub-Arctic boundary layer (mission 19).The aircraft remained at a constant 150-m altitude.