Scientists are studying the mesosphere and lower thermosphere using radar in northern Norway. They found peculiar events with strong upward and downward air movements, happening frequently (up to 2.5 % per month) from 2015 to 2021. Over 700 such events were noted, lasting around 20 min and expanding the studied layer. A total of 17 % of these events had extreme vertical speeds, showing their unique nature.

Investigation of winds is important to understand atmospheric dynamics. In the summer mesosphere and lower thermosphere, there are three main wind flows: the mesospheric westward, the mesopause southward (equatorward), and the lower-thermospheric eastward wind. Combining almost 2 decades of measurements from different radars, we study the trend, their interannual oscillations, and the effects of the geomagnetic activity over these wind maxima.

A rare gravity wave event was observed on the night of 25 April 2017 over northern Germany. An all-sky airglow imager recorded an upward-propagating wave at different altitudes in mesosphere with a prominent wave front above 91 km and faintly observed below. Based on wind and satellite-borne temperature profiles close to the event location, we have found the presence of a leaky thermal duct layer in 85–91 km. The appearance of this duct layer caused the wave amplitudes to diminish below 91 km.

To study long-term trends in the mesosphere and lower thermosphere (70–100 km), we established two summer length definitions and analyzed the variability over the years (2004–2020). After the analysis, we found significant trends in the summer beginning of one definition. Furthermore, we were able to extend one of the time series up to 31 years and obtained evidence of non-uniform trends and periodicities similar to those known for the quasi-biennial oscillation and El Niño–Southern Oscillation.

We introduce a new way of estimating winds in the upper atmosphere (about 80 to 100 km in altitude) from the observed Doppler shift of meteor trails using a statistical method called Gaussian process regression. Wind estimates and, critically, the uncertainty of those estimates can be evaluated smoothly (i.e., not gridded) in space and time. The effective resolution is set by provided parameters, which are limited in practice by the number density of the observed meteors.