Retrieving global cloud top phase patterns: The reliability of spaceborne retrievals

Clouds between -42°C and 0°C may be in a phase of pure-liquid, pure-ice, or a mixture of both. The cloud’s phase affects its radiative effect. However, the cloud phase variability is poorly understood, resulting in a high uncertainty in the climate warming projected by models. Passive retrievals from space have the potential for resolving the temporal variability of cloud phase globally and could help to better understand the drivers of cloud glaciation. In our study, we show that daily passive retrievals of cloud phase are highly unreliable, but the monthly phase frequency for thick stratiform clouds offers a reliable constrain for cloud phase variability.

We find that for both the daily cloud phase and the monthly phase frequency, the agreement among different retrieval algorithms increase for thick stratiform clouds. Moreover, for these clouds, the hemispheric and seasonal contrast of cloud phase agrees better with previous estimations from active retrievals. Using thick stratiform clouds observed from space during 35 years, we offer for the first time an estimation of the seasonal cloud phase contrast resolved globally. This estimation will help increase our understanding of the yet poorly-known drivers of cloud glaciation. Moreover, we offer a previously missing benchmark for climate models to constrain the physical processes behind cloud glaciation in order to reduce the associated uncertainty in climate projections.