Solar Variability and Trend Effects in Mesospheric Ice Layers

Abstract

In this paper we summarize results from the SOLEIL project (SOLar variability and trend Effects in Ice Layers) which was part of the CAWSES priority program in Germany. We present results from LIMA/ICE which is a global circulation model concentrating on ice clouds (NLC, noctilucent clouds) in the summer mesopause region. LIMA/ICE adapts to ECMWF data in the lower atmosphere which produces significant short term and year-to-year variability. The mean ice cloud parameters derived from LIMA/ICE generally agree with observations. The formation, transport, and sublimation of ice particles causes a significant redistribution of water vapor (‘freeze drying’). Model results are now available for all years since 1961 for various scenarios, e.g., with and without greenhouse gas increase etc. Temperatures and water vapor are affected by solar activity. In general it is warmer during solar maximum, but there is a small height region around the mesopause where it is colder. This complicates the prediction of solar cycle effects on ice layers. The magnitude of the solar cycle effect is ∼1–3 K which is similar to the year-to-year variability. Therefore, only a moderate solar cycle signal is observed in temperatures and in ice layers. Temperature trends at NLC altitudes are partly caused by stratospheric trends (‘shrinking effect’). Trends are generally negative, but are positive in the mesopause region. Again, this complicates a simple prediction of temperature trends on ice layers and requires a complex model like LIMA/ICE. Trends in CO2 and stratospheric O3 enhance mesospheric temperature trends but have comparatively small effects in the ice regime. Comparison of contemporary and historic observations of NLC altitudes leads to negligible temperature trends at NLC altitudes (∼83 km). For the time period of satellite measurements (1979–2009) LIMA/ICE predicts trends in ice cloud brightness and occurrence rates, consistent with observations. Temperature trends are not uniform in time but are stronger until the mid 1990s, and weaker thereafter. This change is presumably related to stratospheric ozone recovery. The accidental coincidence of lowest temperatures and solar cycle minimum in the mid 1990s led to large NLC activity. It is important to consider the time period and the height range when studying temperature and ice cloud trends. In the mesosphere temperature trends can be as large as −(3–5) K/decade (in agreement with observations) or rather small, depending on the time period and height range.