Analysis of Arctic ice cloud properties using in-situ and remote sensing measurements

Sammanfattning: Cirrus clouds play an important role in the radiation balance of the atmosphere as theycan have a warming and cooling effect. The resulting net radiation effect depends ontheir micro- and macrophysical properties such as particle size, shape, and numberconcentration. The net warming or cooling effect of clouds is still one of the biggest uncertainties, for example in climate models. For weather and climate models and remotesensing retrievals, precise knowledge of micro- and macrophysical cloud propertiesis therefore necessary. This is true in particular for Arctic cirrus clouds, where we still lack data and need better understanding. Yet, climate change affects high latitudes stronger than other regions. Thus, more knowledge about micro- and macrophysicalproperties of Arctic cirrus clouds is needed urgently.The focus of this thesis is on the retrieval of physical particle properties of Arcticice clouds. Balloon-borne in-situ measurements with concurrent lidar measurementshave been performedinKiruna in winter. The Balloon-borne IceCloud Particle Imager(B-ICI) takes images of ice particles directly inside the cloud. After recovery, the imagesare analysed to gain information about particle shape, size, area, and number concentrationand to determine the extinction coefficient.Whenever possible, concurrent lidar measurements have supplemented the balloonbornemeasurements. Due to balloon drift, there is a spatial and temporal distance betweenB-ICI and lidar. Hence, both instruments do not sample a cloud at the same timeand place. Taking into account the wind speed, it is possible to determine the time of lidarobservation at which the cloud segment probed by balloon was closest to the lidar. However, cloud homogeneity has to be assumed. The results from B-ICI are comparedto extinction coefficients and depolarization ratios obtained fromlidar measurements. Measurement results from B-ICI and lidar measurements are, despite the spatial andtemporal distance, very similar and thus comparable. Clouds consisting of small andcompact particles have a smaller extinction coefficient and depolarization ratio thanclouds which consist of large, complex-shaped particles.For each cloud that has been measured, the cloud origin, i.e. its formation processis determined with the help of back-trajectory modelling. With that, it can be studiedif particle properties depend on cloud origin. This analysis reveals that particle sizeand shape exhibit strong differences with respect to the formation process. If ice particleshave been formed via the liquid droplet phase, they can grow to large sizes andinto complex shapes. If, however, they have been formed directly from vapour or supercooledsolutions, they are smaller and most often compact in shape. Hence, it ispossible to predict the formation process if size distribution and predominant particleiiishapes are known. Or inversely, size distribution and shapes can be predicted by knowingthe formation process. To account for these differences, a new parametrization forparticle size distribution is given that depends on the formation process.While the cloud formation process is depending on temperature, supersaturation,and updraught speed, it should not depend on latitude. In fact, comparing results fromArctic measurements with measurements from other latitudes similarities are recognizable. For example, the new parametrization for particle size distributions of Arcticcirrus clouds depending on formation processes and an established parametrizationfor midlatitude cirrus depending on particle size are very similar. Thus, this thesis andthe attached papers will not just provide better information about the particle propertiesof Arctic ice clouds, it can also be used to improve weather and climate models forall latitudes.