Wicker, Wolfgang (2020) Modelling the upper-tropospheric rotational flow. (Master thesis), Christian Albrechts Univerisität zu Kiel, Kiel, Germany, 79 pp.

Preview

Text Wicker.pdf - Published Version Available under License Creative Commons: Attribution 4.0. Download (10MB) | Preview

Abstract

For this thesis, a fully non-linear barotropic vorticity equation model is modified to be forced
by a time series of realistic Rossby wave source (RWS) diagnosed from the atmospheric
reanalysis product ERA5. Comparing two different model configurations with either weak
or strong linear friction, 41-year long model experiments for the flow at 200hPa and 300hPa
are conducted. In the model, the climatological upper-tropospheric rotational flow is driven
by specified RWS and simulated eddy fluxes that result from stirring. Specifically, the model
is intended as a framework to analyse equivalent barotropic low-frequency variability, i.e.
teleconnections.
However, the expectation to reproduce the upper-tropospheric rotational flow from ERA5
cannot be realised. It is found that the mean-flow westerly jets are too weak and the variance
is too high. In terms of variability, the model performs very differently when viewed from
a global or regional perspective: There is reasonable temporal coherence between model
and reanalysis for area-averaged kinetic energies but the performance for a regional climate
index is low. The strong variance in the model can be attributed to barotropic instabil-
ity. By averaging across an ensemble of model realisations initialised from ten different
initial conditions, this internal variability can be removed. The low-frequency flow field,
however, remains deteriorated compared to reanalysis. It is hypothesised that the model
error arises from the lack of a “baroclinic governor” which bears analogy to the barotropic
governor where baroclinic instability is suppressed in the presence of a small barotropic
shear. This hypothesis is analysed in terms of normal mode solutions to a simple two-layer
quasi-geostrophic model, though not conclusively.
Furthermore, it is found that the traditional RWS is incomplete. Permitting horizontal
divergence but neglecting vertical advection does not allow the horizontal momentum equa-
tions to be written in flux form due to an inconsistency with the continuity equation. A
more complete RWS needs to include the curl of vertical momentum advection: the sum of
vortex tilting (−ωxvp + ωyup) and vertical vorticity advection (−ωζp). These new terms are
particularly important for equatorial dynamics at 200hPa and for the subtropics at 300hPa.

Actions (login required)

View Item