Clouds play a major role in real-world applications like weather forecasting or climate modelling. The climate system is strongly influenced by cloud processes by additionally coupling the key processes (e.g., dynamical, hydrological, boundary-layer, radiation & chemical, and ocean & land processes) with each other. Understanding the role of the microphysics of the cloud particles such as droplets and ice particles plays a key role to understand the influence of clouds on the synoptic-scale circulation (redistribution of sensible and latent heat and momentum), on the radiative forcing (reflection, absorption, and emission of radiation), on the precipitation, and the modification of radiation and planetary boundary layer processes. The work of this thesis aims to help to understand the involvement of a surface-cloud interaction as a role for ice concentrations that surpass modelled ice number concentrations using only the concentration of ice nucleating particles (INPs). The difference of up to 103 L-1 cannot be explained by using common secondary ice mechanisms like the Hallett-Mossop process and also not by increasing the number concentration of INPs in the model. Measurements on top of the Jungfraujoch with an Eddy Flux Measuring system supports the hypothesis of a positive ice crystal flux, generated on the surface that was able to raise the number concentrations to the level of the measurements. Periods of positive and negative fluxes have been measured and analysed. Blowing snow could be ruled out as the wind speed was lower than the needed threshold and through a missing correlation of wind speed and ice number flux. Pictures of ice particles during periods of high fluxes suggest that the possible source of the ice particles is located too far away from the point of measurement to detect a dominant habit, associated with a surface growth. The results show that more experiments closer to the source (e.g., the snow-covered glacier to the South) are needed to learn more about the processes involved. Remote sensing techniques (e.g., radar, satellites) provide a continuous and cheap possibility to obtain information about clouds. As they strongly rely on parameterizations to interpret their products, reliable parallel in-situ and remote measurements are. Therefore, two flights of a recent campaign have been analysed for three different temperature regions to interrogate the disputable high number concentration of ice at small sizes that can be seen in well-established cloud probes. For the comparisons, a new holographic approach of measuring ice number concentration was used. The results show an overestimation of ice number concentration for small sizes up to the order of two in ice clouds. Droplets in the measurements dominate the shape of the size distribution, therefore this overestimation cannot be seen. Hence, much care should be taken in interpreting the results of past and future measurements if used for interpreting remotely measured data (e.g., radar). In a laboratory experiment, the results of the holographic measurements reveal a consistent sizing of particles (i.e., droplets from a droplet gun) inside its sample volume. This emphasises the results from the comparison of the ice number concentrations of the campaign.