8.08 REAL-TIME VERIFICATION OF MESOSCALE MODEL PRODUCTS OVER COMPLEX TERRAIN Manfred Dorninger*, R. Steinacker, T. Gorgas, B. Chimani, S. Tschannett, B. Bica University of Vienna Vienna, Austria 1. INTRODUCTION Verification of NWP model products is done primarily through comparison of model analysed fields or through a comparison of selected surface stations with nearby grid points. The new generation of operational mesoscale forecast models with a horizontal resolution of around 4km does not contain its own data assimilation system in general and consequently does not produce an analysis field. Further, the horizontal resolution is still too rough to compare instantaneous values of surface stations especially over complex terrain with model grid points with success. The analysis tool VERA (Vienna Enhanced Resolution Analysis) fits this gap. It produces surface fields of potential temperature, equivalent potential temperature, mean sea level pressure and 10m wind on a regular basis in real time without using a first guess field. It is therefore best suited to serve as an independent reference for model evaluation. Downscaling of the fields in data sparse areas is performed by using the fingerprint technique, which introduces an a priori physical knowledge of the field distribution over complex terrain. For a detailed discussion on VERA including the treating of the data quality problem see e.g. Steinacker et al. (2000), Häberli et al. (2004), Bica et al. (2005) , Steinacker et al. (2005) and Tschannett et al. (2005). * Corresponding author address: Manfred Dorninger, University of Vienna, Dept. of Meteorology and Geophysics , UZA II, Althanstraße 14, 1090 Vienna, Austria; e-mail: [email protected]2. NWP-MODELS Currently products of three operational NWP- models are compared with the VERA fields. These are the ALADIN, the LM and the IFS-ECMWF model. The evaluation is performed over the larger Alpine area. ALADIN and LM are limited area models with a horizontal resolution of 9 km and 7 km, respectively. Both, the surface fields of the 00 UTC-run as well as of the 12 UTC run are available and evaluated on an hourly interval. The IFS model of the ECMWF is a spectral (T511) and global model. From this model the surface fields of the 12 UTC run on a three hourly interval enter the evaluation procedure 3. COMPARISON METHOD Since the VERA parameters like potential temperature and equivalent potential temperature are not standard model output parameters, they are calculated from the appropriate model parameters of the model surface. In a second step the model data are interpolated to the VERA grid by using a simple inverse distance weighting scheme (Cressman scheme). In a final step the differences between the VERA topography and the model topography are taken into account by using the values of the standard atmosphere for the vertical interpolation. 3. RESULTS 3.1 Case study: Cold front with squall line A squall line forms in front of a cold front over southern Germany on June 23 rd , 2002. The cold air outflow stemming from this squall line rushes down
5
Embed
8.08 REAL-TIME VERIFICATION OF MESOSCALE MODEL PRO … · 8.08 REAL-TIME VERIFICATION OF MESOSCALE MODEL PRO DUCTS OVER COMPLEX TERRAIN Manfred Dorninger *, R. Steinacker, T. Gorgas,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
8.08 REAL-TIME VERIFICATION OF MESOSCALE MODEL PRODUCTS OVER COMPLEX TERRAIN
Manfred Dorninger*, R. Steinacker, T. Gorgas, B. Chimani, S. Tschannett, B. Bica
University of Vienna
Vienna, Austria
1. INTRODUCTION
Verification of NWP model products is done
primarily through comparison of model analysed
fields or through a comparison of selected surface
stations with nearby grid points. The new generation
of operational mesoscale forecast models with a
horizontal resolution of around 4km does not
contain its own data assimilation system in general
and consequently does not produce an analysis
field. Further, the horizontal resolution is still too
rough to compare instantaneous values of surface
stations especially over complex terrain with model
grid points with success.
The analysis tool VERA (Vienna Enhanced
Resolution Analysis) fits this gap. It produces
surface fields of potential temperature, equivalent
potential temperature, mean sea level pressure and
10m wind on a regular basis in real time without
using a first guess field. It is therefore best suited to
serve as an independent reference for model
evaluation. Downscaling of the fields in data sparse
areas is performed by using the fingerprint
technique, which introduces an a priori physical
knowledge of the field distribution over complex
terrain. For a detailed discussion on VERA including
the treating of the data quality problem see e.g.
Steinacker et al. (2000), Häberli et al. (2004), Bica