Universität zu Köln

Kiszler, Theresa ORCID: 0000-0002-2605-1776 (2024). Improving our understanding of cloud phase-partitioning using long-term cloud-resolving simulations of Svalbard. PhD thesis, Universität zu Köln.

PDF Kiszler_Theresa_Dissertation_20231218.pdf Download (11MB)

Abstract

Clouds cover a large part of the Earth and are relevant for the hydrological cycle and the Earth’s radiative budget. Clouds are very frequent in the Arctic, where they warm the surface most of the year in contrast to the rest of the globe. Low-level clouds are especially common in the Arctic and frequently contain supercooled liquid and frozen water simultaneously. The cloud’s composition and radiative properties depend on the aerosols available for droplet or ice particle formation and the microphysical processes that drive the hydrometeor growth and interactions between the hydrometeors. Changes in the composition of clouds due to climate change can affect the cloud’s radiative properties. Such changes can impact the larger climate system and amplify or dampen the warming. However, it is not fully clear what effect the changes in cloud properties and occurrence will have in the future. These uncertainties in models are due to several challenges, and one of the main ones is the representation of subgrid-scale microphysical processes in clouds. These processes, as well as the availability of aerosols, must be parameterised. These microphysical parameterisations inherit the knowledge gaps which still exist related to microphysical processes. With inaccurate parameterisations, models cannot correctly capture the radiative properties of clouds, and climate models can hardly quantify the impact of clouds in a warmer climate. To address this issue, this thesis evaluates cloud-resolving simulations performed for a region of Svalbard, centred at Ny-Ålesund. Ny-Ålesund offers a variety of long-term measurements which are used as a reference. This location further provides a complex topography to challenge the model. As the chosen model was the ICON (ICOsahedral Nonhydrostatic) model, which was developed for Germany, the first goal is to evaluate the model’s overall performance. It is shown that the model can capture large-scale features such as the wind flow and temperature profiles well. Compared to the observations, it also shows a similar cloud occurrence, but it is noticeable that there is an apparent underestimation of the mixed-phase clouds. It was further found that the liquid water path (LWP) is skewed towards higher values, so it became clear that the model glaciated clouds excessively. This misrepresentation of the cloud phase-partitioning has also been found in other models and was therefore further explored in the second study. The second study gave insights into whether the supercooled liquid representation would change using different aerosol settings. For this, the activation and nucleation schemes of vcloud condensation nuclei and ice nucleating particles were adjusted to fit the Arctic better, where the aerosol concentration is lower than in the mid-latitudes. This led to a substantial decrease in water droplets for the lower concentrations and increased raindrop concentrations. The LWP was further found to be better captured compared to the observations for lower LWP values. It seemed, though, that the lower aerosol numbers caused an over-correction of the LWP with no high LWP values for the new setup. Additionally, the change in the number of mixed-phase clouds was minor, indicating that the phase-changing microphysical processes potentially play a larger role in creating mixed-phase regimes. To look further into the microphysical processes, a diagnostic tool was implemented that enables the output of the microphysical process rates. This tool was then used to explore which processes contribute most to the phase changes of hydrometeors. Here, it was found that there were clear differences between polar night and polar day, presumably to a large extent because of the temperature differences. Further, it became clear that phase changes via the vapour phase dominated and this also showed in the occurrence of the Wegener-Bergeron-Findeisen process, which could be quantified. To summarise, three main questions are addressed related to the performance of the ICON model in the Arctic, the impact of aerosols on supercooled liquid, and the microphysical processes contributing to phase changes. Combining all results, the picture emerged that substantial improvements in the representation of the phase-partitioning of clouds, especially low-level clouds, are necessary in ICON. These improvements can only be achieved by looking at both the aerosols and the microphysical processes. A better understanding of which mechanisms in models play a role in determining cloud properties has been achieved through the thorough investigation of long-term cloud-resolving simulations and the cloud microphysical processes. The methodological approach of using a large simulation data set, in combination with the implementation of a diagnostic tool, can be used in future studies and enables an increased understanding of why models behave in certain ways.

Item Type:	Thesis (PhD thesis)
Creators:	Creators Email ORCID ORCID Put Code Kiszler, Theresa UNSPECIFIED orcid.org/0000-0002-2605-1776 UNSPECIFIED
URN:	urn:nbn:de:hbz:38-727228
Date:	2024
Language:	English
Faculty:	Faculty of Mathematics and Natural Sciences
Divisions:	Faculty of Mathematics and Natural Sciences > Department of Geosciences > Institute for Geophysics and Meteorology
Subjects:	Earth sciences
Uncontrolled Keywords:	Keywords Language meteorology English clouds English arctic English simulations English
Date of oral exam:	18 March 2024
Referee:	Name Academic Title Neggers, Roel Prof. Hoose, Corinna Prof.in
Refereed:	Yes
URI:	http://kups.ub.uni-koeln.de/id/eprint/72722