Neske, Stefanie (2017). Towards an improved Turbulence Closure Scheme by analysing ICON Model Simulations. PhD thesis, Universität zu Köln.

[img]
Preview
PDF
thesis_bib.pdf - Accepted Version

Download (7MB)

Abstract

With the progress in supercomputer performance, efforts have been made to per- form global weather prediction simulations at very high resolutions to reduce model uncertainties by explicitly resolving some of the scales which had to be parameter- ized before. This potential leads to newly developed models like ICON (ICOsahedral Nonhydrostatic), but their advantages are accompanied by challenges such that the existing parameterizations and well-established schemes may no longer be adequate. The main aim of this thesis is to analyse whether the well-known Smagorinsky tur- bulence scheme implemented in ICON is scale consistent upon all resolutions and to contribute to an improved closure better suitable for weather and climate prediction with the scale-adaptive grid of the ICON model. To this end, the turbulence scheme is evaluated using ICON model data and further it is examined how subgrid fluxes be- have by using methods like wavelet analysis or a direct comparison between modeled parameters and desired parameters. Two kinds of data sets are used for the analysis. The first one was operated within the framework of the High Definition Clouds and Precipitation [HD(CP)²] project. The simulation period was from 24th of April 2013 until 26th of April 2013, whereas the first and the last day serve as representatives for two different weather situations. Three nested domains with different grid resolutions were selected, analysed separately and compared with each other in order to investigate whether the ICON model output is scale dependent or independent. For the second data set, new simulations have been conducted for the 24th April 2013 with an additional higher resolved nesting domain resulting in four resolutions available for comparison. The analysis of both data sets demonstrates that the turbulence parameterization strongly depends on scale and that the runs on different resolutions substantially dis- agree in their flux estimates. The results suggest that the closure requires improvement in order to exploit all advantages and new possibilities of the ICON model. Different kinds of analysis methods were used with the focus to better understand the behaviour iof the subgrid fluxes. The results point out that eddies of sizes close to the grid scale have the largest influence on the subgrid fluxes. This contradicts the K-theory, which assumes that small eddies have the largest influence on the subgrid flux. Since K- theory is the basis of the closure used in the ICON model and that is also basis of many other closures, this concept has to be questioned. Considering the energy spectrum for turbulence and the grid sizes of the available data sets, points out that the closure probably just captures the smaller scales up to about 200 m grid size. Accordingly, the closure does not seem to be sufficient for most of the resolutions used in this thesis. As a result of this thesis, two hypotheses for an improved turbulence closure are pro- posed. An improved closure directly influences the quality of the ICON model weather and climate prediction capability. Furthermore, it can also be a useful enhancement for other scientists to improve weather prediction models.

Item Type: Thesis (PhD thesis)
Creators:
CreatorsEmailORCIDORCID Put Code
Neske, Stefaniestefanie.neske@gmail.comUNSPECIFIEDUNSPECIFIED
URN: urn:nbn:de:hbz:38-75655
Date: 2017
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: Natural sciences and mathematics
Earth sciences
Uncontrolled Keywords:
KeywordsLanguage
turbulence, closure, ICON, modellingEnglish
Date of oral exam: 24 April 2017
Referee:
NameAcademic Title
Shao, YapingProf. Dr.
Elbern, HendrikPD Dr.
Refereed: Yes
URI: http://kups.ub.uni-koeln.de/id/eprint/7565

Downloads

Downloads per month over past year

Export

Actions (login required)

View Item View Item