NO₂

  

Nadir NO₂ results along the 26 April 1997 flight track are given in Figure 6.23 using 450-500 nm. In order to reduce the amount of noise which arises when taking the ratio of individual spectra, a spatial 1-2-1 running average of both the nadir and horizontal flux spectra was used. As is evident from Figure 6.23a, even after the spatial averaging, the noise is much larger than was observed for the UV ozone. The sharp decreases near 73 ksec corresponds to that observed in the ozone, the result of a decrease in ER-2 altitude down to 15 km. However, soon after another decrease of comparable size is evident although there is no apparent reason for this. This flight was known to have skirted the edge of the polar vortex, which lasted into late April in 1997 (Coy et al., 1997), and this could be evidence of denitrification in the vortex itself or a vortex fragment.

The VCDs, in panel (b), vary between 3-7 $\times 10^{15}$ cm^-2 with an average of about 5 $\times 10^{15}$ cm^-2. This is likely too large as a total VCD of NO₂ has been measured in the Arctic during summer between 1990 and 1995 at 5-7 $\times 10^{15}$ cm^-2 (Goutail et al., 1995; Schmidt, 1997). Similar to ozone, roughly half the total NO₂ column will reside above the ER-2 in the Arctic during spring and summer. The NO₂ dependence on latitude and SZA are shown in Figure 6.23c and 6.23d, respectively. There is no clear latitude dependence and the northern and southern legs do not agree as there seems to be an overall decrease throughout the flight.

The 6 May 1997 flight was also analysed with the results presented in Figure 6.24. VCDs were found to be similar in both magnitude and variability throughout this flight. This apparent overestimation is similar to that observed for the visible ozone which was retrieved in roughly the same spectral region. Again, this could be the result of a problem with the albedo or it could be a systematic problem. Clearly further work is required to determine the cause.