Contrasting ice formation in Arctic clouds: surface-coupled vs. surface-decoupled clouds

In the Arctic summer of 2017 (1 June to 16 July) measurements with the OCEANET-Atmosphere facility were performed during the Polarstern cruise PS106. OCEANET comprises amongst other instruments the multiwavelength polarization lidar PollyXT_OCEANET and for PS106 was complemented with a vertically pointed 35 GHz cloud radar. In the scope of the presented study, the influence of cloud height and surface coupling on the probability of clouds to contain and form ice is investigated. Polarimetric lidar data were used for the detection of the cloud base and the identification of the thermodynamic phase. Both radar and lidar were used to detect cloud top. Radiosonde data were used to derive the thermodynamic structure of the atmosphere and the clouds. The analyzed data set shows a significant impact of the surface-coupling state on the probability of ice formation. Surface-coupled clouds were identified by a quasi-constant potential temperature profile from the surface up to liquid layer base. Within the same minimum cloud temperature range, ice-containing clouds have been observed more frequently than surface-decoupled clouds by a factor of up to 6 (temperature intervals between −7.5 and −5 ∘C, 164 vs. 27 analyzed intervals of 30 min). The frequency of occurrence of surface-coupled ice-containing clouds was found to be 2–3 times higher (e.g., 82 % vs. 35 % between −7.5 and −5 ∘C). These findings provide evidence that above −10 ∘C heterogeneous ice formation in Arctic mixed-phase clouds occurs by a factor of 2–6 more often when the cloud layer is coupled to the surface. In turn, for minimum cloud temperatures below −15 ∘C, the frequency of ice-containing clouds for coupled and decoupled conditions approached the respective curve for the central European site of Leipzig, Germany (51∘ N, 12∘ E). This corroborates the hypothesis that the free-tropospheric ice nucleating particle (INP) reservoir over the Arctic is controlled by continental aerosol. Two sensitivity studies, also using the cloud radar for detection of ice particles and applying a modified coupling state detection, both confirmed the findings, albeit with a lower magnitude. Possible explanations for the observations are discussed by considering recent in situ measurements of INP in the Arctic, of which much higher concentrations were found in the surface-coupled atmosphere in close vicinity to the ice shore.