Increasing concentration of greenhouse gases, particularly of CO₂, in the atmosphere causes well-known heating of the troposphere and surface. However, the increasing concentration of CO₂ also affects higher levels of the atmosphere, the stratosphere, mesosphere, thermosphere, and ionosphere, where it results in remarkable long-term trends. This article reviews significant progress in investigations of long-term trends in the mesosphere, thermosphere, and ionosphere during the period 2018–2022.

This paper is devoted to the occurrence of discontinuities in the ozone concentration data from the selected reanalyses, because they have large impact to the results of trend studies. The discontinuity occurrence is reanalyse dependant. The discontinuities frequently occur at the middle stratosphere and in the troposphere for a certain reanalyses. According our opinion, the reanalyses data can be used in trend studies especially in the lower stratosphere.

The longitudinal structure in geopotential heights and meridional wind is analysed for 1979–2013 in order to find its persistence and altitudinal dependence with focus on anomalous years. Substantial deviations from the average pattern are studied for Januaries – typically the second (Euro-Atlantic) peak extends to much higher altitudes than usual. The decisive role in the existence of anomalous years appears to be played by the stationary planetary wave filtering by the zonal wind pattern.

An unusual geomagnetic storm effect on the ionosphere in the form of localized enhancement of total electron content southward of South Africa is investigated using data from different ionosphere sounding facilities. Global ionospheric maps allow us to get the occurrence rate of such irregularities over the period of 2002–2016 and to see its correlation with solar cycle. The events we detected occur during geomagnetic storms, but not every geomagnetic storm produces such an effect.

A study of trends in the middle stratosphere using comparisons of four main reanalyses (ERA-Interim, JRA-55, MERRA and NCEP/NCAR). We identified that all four reanalyses show very similar trends on a season or monthly basis. We also compute trends for each grid point not as a zonal mean. This approach shows detailed features in the trend studies in both hemispheres.