Climatology of polar mesospheric clouds: 2. Further analysis of solar mesosphere explorer data

Abstract

An analysis is presented of nine summer seasons of Solar Mesosphere Explorer (SME) satellite data used to study large‐scale geographical and temporal variability of polar mesospheric clouds (PMC). PMC are scattering layers occurring in the cold summertime polar cap atmosphere near 85 km. This region is often supersaturated with respect to water‐ice crystals in a narrow height domain where temperatures fall below 140 K. An earlier analysis showed that PMC appear to be a regular feature of the arctic and antarctic mesosphere during summer solstice. PMC grow in brightness with latitude frcm near their detection threshold at about 55° latitude toward the pole. In the current study we examine a larger set of SME data, from 1981 to 1985. In addition we have removed the Rayleigh‐scattering background from the cloud data. We find that this subtraction results in a better defined cloud activity index in the low latitude regions where clouds are infrequent. We find that the PMC season begins at the highest latitudes observed and propagates to lower latitudes within 10 to 20 days. This occurs at about 20 to 40 days before summer solstice. The season ends about 60 days after solstice, with the maximum occurrence frequency to be found at 1 to 3 weeks following solstice. The beginning and ending of the PMC season occur about 1 to 2 weeks later than published values of the noctilucent cloud season, as observed from the ground at northerly latitudes of 50°–65°. The northern PMC extend farther from the pole by about 5° than their southern counterparts. A comparison with rocket data for the seasonal variation of temperature at 70° latitude indicates similar time lags of about 3 weeks after solstice for minimum mesopause temperature and maximum cloud frequency. This suggests that variations of temperature and/or the accompanying upward advective water vapor flux are responsible for the seasonal variations of PMC. We present observational differences between PMC and noctilucent clouds as well as a possible unification of our understanding of the two phenomena. A study of the longitudinal variability of PMC indicates no significant large‐scale variations; variations of the scale of 40° are masked by the uneven sampling from year to year. There is significant interannual variability (of the order of 15%) of unknown cause, but no consistent year‐to‐year trends in activity. There is no indication of an influence of the location of the auroral oval on the geographical distribution of PMC; any possible causal connection between PMC and auroral activity remains unknown.