Temporal variability of lower stratosphere temperature

Abstract

Inter-monthly to inter-decadal global variability of lower stratosphere temperature (LST) is studied in order to improve current knowledge on its variability and trends, as well as natural and anthropogenic influences upon it. Principal Component Analysis (PCA) with S-mode Varimax rotated PCA were used. The first seven components, which explain 70% of variance make it possible to determine homogeneous LST behaviour zones with little overlap between areas, and practically no unclassified areas. Composite time series, referred to as reference series, in the core of the subregions defined by each of the PCs, were calculated in order to obtain the temporal patterns. The equatorial-tropical zone and the subtropical area display warmings caused by the eruptions of El Chichon and Mt. Pinatubo volcanoes as well as the strong influence of the Quasi-Biennial Oscillation (QBO) which leads to equatorial warming (cooling) in the west (east) phase and cooling (warming) in subtropical latitudes. Only low latitudes show some kind of global teleconnection between hemispheres. Significant correlation with several ocean/atmosphere index time-series like ENSO, Antarctic and Arctic Oscillations (AAO, AO), Arctic Circumpolar Vortex was detected over latitudinally separate regions. Antarctic and Arctic ozone hole values were contrasted with warming and cooling features registered in mid and high latitudes in both hemispheres. The LST reference series exhibit a negative trend, commonly attributed to the increase in greenhouse gases that lead to a warming of the troposphere and a cooling of the stratosphere, in all sub regions. The highest cooling rate of -0.65°C/decade is detected in the Gobi desert, and the lowest values of -0.1 °C/decade over the NE of Canada and Greenland which indicates the great longitudinal variability that the LST trends may present. The difference with other authors is mainly due to the fact that results are based either on latitudinal averages or radiosonde data. © StudiaGeo s.r.o. 2005.