The October Effect in the Neutral Atmosphere Simulated by Different General Circulation Models

Kurzfassung

The October effect is well known as a sharp decrease in the amplitude of Very Low Frequency (VLF) radio waves reflected in the D-region (60–90 km). However, the underlying mechanism of the October effect remains unclear. Recent studies suggest that neutral atmosphere dynamics may play a key role. Coincident with the October effect in the ionized D-region, a warming occurs in the lower mesosphere. We refer to this feature as the neutral October effect. Notably, this warming is absent in spring, resulting in a pronounced spring-fall asymmetry. This spring-fall asymmetry is reproduced by models that assimilate satellite observations in the mesosphere, such as MERRA-2 and JAWARA. However, MERRA-2 captures the neutral October effect only after 2005, when mesospheric satellite data began to be assimilated. In contrast, models that are free-running in the mesosphere generally fail to reproduce the neutral October effect (e.g., WACCM-X, ERA5, GAIA, UA-ICON). One notable exception is CMAM30, which successfully reproduces the neutral October effect. This study aims to investigate the mechanism responsible for the neutral October effect in the mesosphere and to identify why most models struggle to simulate it. Using CMAM30 model output and satellite observations from SABER and MLS, we examine the dynamics of the fall transition, with a particular focus on planetary and gravity waves. Our results indicate that gravity waves play a key role in generating the neutral October effect. This finding provides further evidence of deficiencies in current gravity wave parameterizations. At present, these deficiencies can be mitigated only by assimilating mesospheric observations or by increasing model resolution, as demonstrated by a high-resolution WACCM-X simulation.