The reactive nitrogen (NOy), total water, and particle components of the polar stratospheric cloud (PSC) aerosol in the Arctic are studied using in situ aircraft measurements in the lower stratosphere. The results are compared to findings from the Antarctic derived using similar measurements and interpretive techniques. The Arctic data show that particle volume well above background values is present at temperatures above the frostpoint, confirming the result from the Antarctic that the observed PSC's are not water ice particles. NOy measurements inside a PSC are enhanced above ambient values consistent with anisokinetic sampling of particles containing NOy. Furthermore, relative changes in the measured particle volume along a flight track are well correlated with changes in the amount of NOy estimated to be in the particle phase. With the exception of data from one flight, assuming that the composition of the PSCA particles is nitric acid trihydrate (NAT), the HNO3 content of the measured particle volume is consistent with the amount of HNO3 predicted to be available for condensation. To this extent, the Arctic observations are consistent with NAT as the composition of PSC particles, in agreement with the Antarctic findings.

In the Arctic data over long segments of several flights, calculations show saturation with respect to NAT without significant PSC particle growth above background. PSCs in the Arctic are not observed in situ until the apparent saturation ratio of HNO3 with respect to NAT is greater than 10, in marked contrast to the Antarctic, where PSCs are observed in conditions of apparent HNO3 saturation of 1 and above. This difference cannot be resolved by known measurement uncertainties. Also, a discrepancy is noted in the comparison of the amount of condensed HNO3 derived from the particle distribution measurements with that derived from the NOy measurements, assuming a NAT composition for the particles.Although the relative variations in the derived quantities are similar, as in the Antarctic, the mean values consistently disagree by about a factor of 2. The differences suggest that there may be systematic errors in the data and/or physical assumptions used in the analysis. Several possibilities are discussed. ¿ American Geophysical Union 1992