To better understand possible reasons for the diverse modeling results and large discrepancies of the detected solar fingerprints, we took one step back and assessed the “initial” solar signals in the middle atmosphere based on a set of ensemble historical simulations with multiple climate models – the Flexible Ocean Climate Infrastructure (FOCI), the ECHAM/MESSy Atmospheric Chemistry (EMAC), and the Max Planck Institute for Meteorology Earth System Model in high-resolution configuration (MPI-ESM-HR). Consistent with previous work, we find that the 11-year solar cycle signals in the shortwave heating rate (SWHR) and ozone anomalies are robust and statistically significant in all three models. These initial solar cycle signals in the SWHR, ozone, and temperature anomalies are sensitive to the strength of the solar forcing. Correlation coefficients of the solar cycle with the SWHR, ozone, and temperature anomalies linearly increase along with the enhancement of the solar cycle amplitude. This reliance becomes more complex when the solar cycle amplitude – indicated by the standard deviation of the December–January–February mean F10.7 – is larger than 40. In addition, the cold bias in the tropical stratopause of EMAC dampens the subsequent results of the initial solar signal. The warm pole bias in MPI-ESM-HR leads to a weak polar night jet (PNJ), which may limit the top-down propagation of the initial solar signal. Although FOCI simulated a so-called top-down response as revealed in previous studies in a period with large solar cycle amplitudes, its warm bias in the tropical upper stratosphere results in a positive bias in PNJ and can lead to a “reversed” response in some extreme cases. We suggest a careful interpretation of the single model result and further re-examination of the solar signal based on more climate models.
