Comparing interhemispheric differences of mesosphere/lower thermosphere dynamics from ground-based observations and three general circulation models

Abstract

Meteor radars have been proven to be valuable assets in investigating and monitoring mesosphere/lower thermosphere winds for the last two decades. In this study we present a comparison of almost continuous meteor radar measurements obtained from six meteor radars located at mid- and polar conjugate latitudes in both hemispheres. For this purpose we havecompiled harmonized data sets for the Sodankylä (67.9°N, 21.1°E), Esrange (67.4°N, 26.6°E), Davis (68.6°S, 78.0°E), Collm (51.3°N, 13.0°E), Tierra del Fuego meteor radar (53.7°S, 67.7°W) and the Canadian Meteor Orbit Radar (CMOR) (43.3°N, 80.8°W). The analysis revealed characteristic differences between the northern and southern hemisphere in the mean winds, in the strength of the mesospheric jets as well as in the tidal climatologies. In particular, semidiurnal tides show significant and distinct interhemispheric differences, notably a strong seasonal asymmetry in amplitude and phase, most prominent during the hemispheric fall transition from September to November. We also compared the observational climatologies with predictions from the three general circulation models GAIA, WACCM-X(SD) and ICON-UA. The model data were analyzed by simulating the radar in the model domain and applying an identical diagnostic to extract mean winds, tides and gravity wave activity. Our comparison reveals substantial differences between model and observational mean winds and tides that vary seasonally, by model and hemisphere. GAIA indicates similar winds during the hemispheric winter conditions compared to the observations, whereas WACCM-X(SD) showed a better agreement to the observations for the summer zonal wind reversal. The models are only partially able to capture interhemispheric differences, with the free-running ICON-UA model best reproducing the interhemispheric difference of the semidiurnal tide in reasonable agreement to observations.