High-resolution ensemble forecasts of a polar low by non-hydrostatic downscaling
A short range, limited area ensemble prediction system, LAMEPS, is currently in operational use at the Norwegian Meteorological Institute. It employs 3D-Var for 6 hourly data assimilation cycling for analysis of the control forecast. Initial time and lateral boundaries ensemble perturbations are com...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Master Thesis |
| Language: | English |
| Published: |
2009
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10852/12619 http://urn.nb.no/URN:NBN:no-23294 |
| Summary: | A short range, limited area ensemble prediction system, LAMEPS, is currently in operational use at the Norwegian Meteorological Institute. It employs 3D-Var for 6 hourly data assimilation cycling for analysis of the control forecast. Initial time and lateral boundaries ensemble perturbations are computed from the 20 + 1 member TEPS (targeted EPS at ECMWF). LAMEPS is run with the quasi-hydrostatic model HIRLAM version 7.1.4 on a 12 km horizontal grid mesh. In this study we have downscaled each LAMEPS member with the non-hydrostatic UK Met Office Unified Model (UM) version 6.1 in order to study the predictability and the predictions of extreme weather related to a polar low observed in the Barents and Norwegian Seas between 3 and 4 March 2008. This event was extensively covered by the observation campaign of the IPY-THORPEX project. UM is in this study configured with 4 km horizontal grid mesh. The domain size has been investigated by using two different domains, one with 390 x 490 and one with 300 x 300 grid points. Furthermore, the sensitivity to the physical parameterization in the stable boundary layer has also been explored. Regular observation data, satellite data, and IPY-THORPEX campaign data have been used to compare with the ensemble forecasts. Probabilities of different meteorological parameters and occurrence of extreme weather events have been studied along with ensemble means, ensemble spread and control runs. In addition, two new model diagnostics for comparing against observation data have been developed. These are cloud top temperatures and tracking of the polar lows path. The ensemble forecast shows clear improvements by increasing horizontal resolution with non-hydrostatic dynamics. However, the size of the integration domain affects the prediction substantially. The improvements are greatest for the large domain. The forecasts are also sensitive to the physical parameterization. The experiments with less vertical mixing in the stable boundary layer reduce the area of high probability for the large domain. The results of the tracking algorithm, which finds the strongest mesoscale track in each ensemble member, show that the location of the strongest track depends on domain size and the perturbation of the physics. |
|---|