Langguth, Michael: Representation of deep convection at gray-zone resolutions - Implementing and testing the HYbrid MAss flux Convection Scheme (HYMACS) in the ICON model. - Bonn, 2022. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc:
author = {{Michael Langguth}},
title = {Representation of deep convection at gray-zone resolutions - Implementing and testing the HYbrid MAss flux Convection Scheme (HYMACS) in the ICON model},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2022,
month = jul,

volume = 94,
note = {Despite the increasing computing resources, the grids of contemporary atmospheric models are often still too coarse to explicitly represent convective processes. Since these processes are known to be an important driver of atmospheric dynamics, convection parametrization schemes must be deployed which have been developed over several decades. However, recent applications of regional climate and operational numerical weather prediction models have reached spatial resolutions where the overturning circulation of deep convection becomes partly resolved onto grid-scale. The so-called gray-zone of deep convection imposes challenges to the conventional parametrization approach which have attained increasing scientific attention in the last two decades.
The present work implements and tests extensively the HYbrid MAss flux Convection Scheme (HYMACS) in the ICOsahedral Non-hydrostatic (ICON) model. In contrast to other convection parametrization schemes, HYMACS passes the compensational subsidence to the grid-scale dynamics, thereby allowing for a net mass transport. While the scheme has been developed since the pioneering work of Kuell et al. (2007), it has only been tested in a couple of case studies with the COSMO (COnsortium for Small-scale Modeling) model. Although the hybrid scheme improved the representation of convection at gray-zone resolutions in these studies, a statistically well founded assessment of the merits of HYMACS is still outstanding. Besides, its implementation into ICON is appealing since the new hosting model is designed to operate over a broad range of spatial resolutions.
This thesis starts with a in-depth introduction of the theoretical framework of HYMACS and documents recent developments of the scheme. Apart from some required adaptions of the physics-dynamics coupling with ICON, problems in conjunction with the numerical filter in the model’s dynamical core are identified. The operational anisotropic divergence damping operator distorts the dynamical flow response to a parametrized net mass transport and therefore has to be revised. Different numerical filter operators are investigated in dynamical core tests on the sphere and in mass lifting experiments. Based on these tests, a revised filter configuration is proposed which is compatible with HYMACS and which efficiently removes computational noise.
With the revised numerical filter configuration, a series of re-forecasts over Central Europe spanning a summery three-monthly period is conducted to analyze the performance of HYMACS in ICON. It is demonstrated that the hybrid scheme captures the convectively driven diurnal cycle of precipitation better than the operational convection parametrization scheme. The modelled marginal distribution of precipitation amounts and the spatial patterns of precipitation also get improved. Albeit the statistical analysis confirms the results of former case studies, issues to the net mass transport of shallow convection are identified as well. Nonetheless, the merits are encouraging and this work is considered to serve as the basis for further developments focusing on the scale adaptivity of HYMACS in the modeling framework of ICON.},

url = {}

The following license files are associated with this item: