A problem with determining the change from observed data is that there happend large changes of the aerosol optical depth with a significant decrease sind about 2010. Therefore, we examine the difference between the JFMA-priod in 2020 and in 2019/18.

A number of location were chosen (see map), with Wuhan and Beijing in the center of the lockdown, a unaffected remote station in the far west, Lanzhou, and two stations in the south, Guilin and Lishui.

The relative warming shows a marked diurnal cycle, with maxima at local early afternoon. The effect amounts to about 0.7K at Wuhan and Beijing, is weaker in the southern stations and absent in the far west.

Differences of 2020  and 1990-2019 daily mean temperatures according to the JRA55 analysis - relative to the same difference in Wuhan.
Thus, in the surrounding area of Wuhan the warming of 2020 is larger than in more remote areas, indicating that the local warming in Wuhan is not due to regional synoptic processes.

To further estiimate the expected climatic  reponse to an increased aerosol optical depth due to the lockdown an ensemble of simulations covering China extending across January to April 2020 was done.

The model is COSMO._CLM with a horizontal grid of about 0.22^o , with lateral forcing given by .JRA55. The radiation scheme uses the aerosol optical depth, single scattering albedo and asymmetry factor given by the MAC-v2 climatology. Three pairs of simulations were done, one with the anthropogenic load given in 2019 (control) and one with the anthropogenic load everywhere reduced by 50%. This is thought to be a rough estimate of the changed conditions in 2020 due to the lockdown

The effect causes in Beijing and Wuhan a mean increase of temperature onf 0.5K

Ensemble-mean for Feburary 2020 reprenting the effect of a50% reduction of anthrogenic AOD in all fo China. Spatially, the whole central and southern part of china is affected by an increase of temperature.

These effects repeat in all three pairs of simulations.

• At selected central China stations temperature were found to be higher than in previous 2 years. This warming goes with a marked diurnal signal. Maximum warming in the early afternoon (06 UTC), weakest at night (18 UTC)
• This may be related to a general warming of large swaths of Northern Asia  which is not related to local aerosol forcing. Indeed, also the stations outside the immediate strong lockdown are showing an albeit weaker warming. Thus, the difference 2020 minus 2919/2018 may overestimate the lockdown effect.
• The ad-hoc numerical experiment with an overall 50%-reduction of anthropogenic aerosol load results in an overall warming is of the order of half a degree with maxima at about noon time, and minima during night.
• The numerical experiment indicates that the reduction of the atmospheric optical depth does not lead to phases of larger deviations of local time series in the simulations has hardly an effect on synoptic dynamics, and the temperature increase is a local effect, albeit everywhere.