Sodankylä Geophysical Observatory, Tähteläntie 112, FIN-99600 Sodankylä, Finland
In 1988, it was estimated by a two-dimensional numerical model, that the stratopause cools by up to 15 K if the concentrations of greenhouse gases are doubled . Later, a global mean model predicted the respective effect in the upper atmosphere to be a cooling of the mesosphere and thermosphere of 10 K and 50 K, respectively . Many papers have been published thereafter trying to reveal long-term trends in a number of ionospheric parameters which could be due to an enhanced greenhouse effect. Besides temperature measurements by sounding rockets and lidars, the altitude variation of the layer of maximum electron density seems to be one of the most promising indirect approaches : the layer is suggested to lower down due to shrinking of the underlying cooled atmosphere. The so-called F2-layer peak, the height of which is denoted by hmF2, is monitored by ionosondes since the 1930s. The F2-layer peak height can be estimated by means of empirical formulae from the maximum usable frequency factor and the critical frequencies of the E and F2 layer. A number of authors have observed a lowering of the F2 layer, e.g. at Sodankylä , and speculatively connected their results to greenhouse cooling. However, recently down as well as up trends were found in studies of data of many ionosondes world-wide .
Time series of F2-peak height are strongly correlated with solar activity, but also exhibit signs of geomagnetic activity variations as well as annual and semi-annual variations . A possible linear trend can be extracted by fitting a multi-parameter model to the time series. The choice of the model is somewhat arbitrary, since there are, e.g., different representations of solar activity.
Here I will show data of many stations around the world and discuss how the magnitude of the long-term trend is influenced by the choice of the empirical formula for hmF2. Furthermore, different multi-parameter models will be fitted to time series of F2-peak heights and the influence of the choice of the model on the trend magnitude will be discussed.
 G. Brasseur and M. H. Hitchman, Science 240 (1988) 634.
 R. G. Roble and R. E. Dickinson, Geophys. Res. Lett. 16 (1989) 1441.
 H. Rishbeth, Planet. Space Sci. 240 (1990) 945.
 Th. Ulich and E. Turunen, Geophys. Res. Lett. 24 (1997) 1103.
 H. O. Upadhyay and K. K. Mahajan, Geophys. Res. Lett. 25 (1998) 3375.
 H. Rishbeth, K. J. F. Sedgemore-Schulthess, Th. Ulich, Ann. Geophys. 18 (2000), in press.