Deep mixed-phase clouds pose an observational challenge to ground-based remote sensing.

Since both the amount and vertical distribution of cloud liquid water strongly influence cloud radiative effects, cloud precipitation formation, and consequently cloud lifetime, improved estimates of these properties are needed to better represent deep mixed-phase clouds in atmospheric models.

In this talk I will show how the information content of cloud radar Doppler spectra can be exploited to learn more about the composition of deep mixed-phase clouds. Firstly, I will introduce a cloud-radar based machine learning approach for detecting supercooled liquid layers in deep or multilayer clouds. The presence of supercooled liquid in mixed-phase clouds is a prerequisite for riming as precipitation formation mechanism. So secondly, I will show how riming can be detected using cloud radar Doppler spectra. I will conclude the talk with an outlook on an upcoming project in mountainous terrain where the introduced techniques will be applied.