 |
| Titel |
A Case-study in Forecasting Low-lying Inversions and Stratocumulus Cloud Cover in the Bay of Biscay |
| VerfasserIn |
Anna Possner, Elias Zubler, Oliver Fuhrer, Ulrike Lohmann, Christoph Schär |
| Konferenz |
EGU General Assembly 2013
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 15 (2013) |
| Datensatznummer |
250084867
|
|
|
|
|
|
| Zusammenfassung |
| Many regional operational forecasting models struggle to simulate low-lying strong
temperature inversions. In order to understand and begin to improve on this apparent deficit,
we investigate the processes of inversion formation and their implementation. Next to the
large-scale subsidence, the turbulent vertical mixing is a key dynamical process in the
formation of inversions. In this case study we use the COSMO model, where turbulence is
parameterised by the Mellor and Yamada (1974,1982) scheme, which is a state-of-the-art
turbulence scheme used by many limited-area weather prediction and regional as well as
global climate models. Here we investigate the sensitivities with respect to two key
parameters (a prescribed minimum threshold for the vertical mixing and the maximum length
scale of turbulent eddies) on forecasting shallow, cloud-topped marine boundary
layers (MBLs) in the Bay of Biscay for January 2003. This particular period was
characterised by an initial 10 – 12 K strong temperature inversion at a MBL height between
500 – 800 m with stratocumulus cloud cover below, which was slowly lifted by an
impeding cold front. Furthermore we tested how well the model could form as
well as maintain the inversion in 2-km and 12-km simulations with varied lead
times of 12h – 72h and how the vertical cloud distribution was affected. The cloud
microphysics was represented by the 2-moment scheme of Seifert and Beheng
(2006), providing prognostic equations for cloud water as well as droplet number
concentration and hence a cloud-droplet size distribution. We used the 2-moment
scheme, as it leads to a more realistic representation of cloud drop activation and
precipitation formation by autoconversion, which is crucial for maintaining the
stratocumulus layer. Additionally, the 2-moment scheme will be needed for future work
focusing on ship tracks, which were also observed during the period analysed in this
study.
The key findings of this study showed that limiting the vertical mixing, as it is currently
done in COSMO as well as other regional models, to at least 1m²/s considerably weakens the
inversion and produces a much too dry MBL with little cloud cover. On the other hand little
sensitivity was observed towards the prescribed maximum length scale of the turbulent
eddies. Additionally, the pre-frontal lifting was found to increase with simulation time
indicating stronger frontogenesis. |
|
|
|
|
|
|