The concentration of tropospheric carbon dioxide at Mauna Loa, Hawaii, has increased by about 13.5% between 1958 and 1994. Based on the assumption of doubled greenhouse gases, numerical models predict the stratopause to cool by up to 15 K. The mesosphere and thermosphere are expected to cool by 10 K and up to 90 K, respectively.
Evidence of this "greenhouse cooling" is limited. Temperature measurements from satellites and by lidar show cooling trends. However, these time series are still too short in order to draw definite conclusions. Decades of temperature measurements by sounding rockets suffer from their irregular distribution in time and location, but seem to agree that the atmosphere between 25 km and 75 km altitude has cooled slightly. The absorption of cosmic radio noise in the ionospheric D region is suggested to decrease due to mesospheric cooling. Model calculations tentatively support this idea. The altitude of maximum electron density is expected to lower due to shrinking of the underlying cooling atmosphere. Ionosonde data from all over the world have been analysed and show negative as well as positive trends.
Here I review long-term trends in the upper atmosphere as observed in various data sets and compare them with each other and with the respective predictions. I will present long-term data from the Finnish Riometer Chain as well as from the EISCAT incoherent scatter radars in Northern Fenno-Scandia.