VISIBLE AND NEAR-ULTRAVIOLET SPECTROSCOPY AT MCMURDO STATION, ANTARCTICA .9. OBSERVATIONS OF OCLO FROM APRIL TO OCTOBER 1991

The first spectroscopic measurements of chlorine dioxide throughout An tarctic fall, winter, and spring were carried out at McMurdo Station d uring April to October 1991. Two different observing modes were employ ed to extend the measurements over the broadest possible seasonal rang e: direct Moon measurements were used as well as observations of the s cattered light from the sky at an 80-degrees angle relative to the zen ith in the direction toward the Sun. The latter observing scheme facil itates measurements during the late twilight, when the brightness in t he more conventional zenith viewing direction is much weaker. The meas urements made using both the scattered light and the lunar sources sho w that OClO was below detection limits in late April and May (twilight OClO vertical column <3x10(12) cm-2), for a corresponding total ozone loss mte due to the coupling of chlorine and bromine of no more than about 0.25 DU/d. OClO was first observed in late June using lunar spec tra, at which time high nighttime vertical column abundances of about 1.0x10(14) cm-2 were measured, implying extensive conversion of inorga nic chlorine to reactive forms in the lower stratosphere. The observat ion of OClO in midwinter implies that the air parcels had been exposed to sunlight, presumably via wave-driven excursions of airflow to lati tudes outside the polar night region. High levels of OClO were measure d using moonlight near the full Moon phases in late July, August, and September. Further, these measurements of high OClO indicate that sign ificant ozone loss should have taken place during midwinter. The winte r ozone loss rate suggested by these data is likely to increase in the future as atmospheric loading of chlorine and bromine continues to ri se. In addition, scattered light measurements in late July and August revealed very large twilight column abundances then, implying signific ant ozone loss rates. Somewhat smaller twilight column abundances were measured in September, and the OClO dropped below detection levels in October.