Arctic clouds in ECHAM6 and their sensitivity to cloud microphysics and surface fluxes
Compared to other climate models, the MPI-ESM/ECHAM6 is one of the few models that is able to realistically simulate the typical two-state radiative structure of the Arctic boundary layer and also is able to sustain liquid water at low temperatures as is often observed in high latitudes. To identify processes in the model that are responsible for the abovementioned features, we compare cloud properties from ECHAM6 to observations from CALIPSO-GOCCP using the COSP satellite simulator and perform sensitivity runs. The comparison shows that the model is able to reproduce the spatial distribution and cloud amount in the Arctic to some extent but a positive bias in cloud fraction is found in high latitudes, which is related to an overestimation of low- and high-level clouds. We mainly focus on low-level clouds and show that the overestimated cloud amount is connected to surfaces that are covered with snow or ice and is mainly caused by an overestimation of liquid-containing clouds. The overestimated amount of Arctic low-level liquid clouds can be related to insufficient efficiency of the Wegener–Bergeron–Findeisen (WBF) process but revising this process alone is not sufficient to improve cloud phase on a global scale as it also introduces a negative bias over oceanic regions in high latitudes. Additionally, this measure transformed the positive bias in low-level liquid clouds into a positive bias of low-level ice clouds, keeping the amount of low-level clouds almost unchanged. To avoid this spurious increase in ice clouds, we allowed for supersaturation with respect to ice using a temperature-weighted scheme for saturation vapor pressure but this measure, together with a more effective WBF process, might already be too efficient at removing clouds as it introduces a negative cloud cover bias. We additionally explored the sensitivity of low-level cloud cover to the strength of surface heat fluxes; by increasing surface mixing, the observed cloud cover and cloud phase bias could also be reduced. As ECHAM6 already mixes too strongly in the Arctic regions, it is questionable if one can physically justify it to increase mixing even further.