Note: all ALATNET related R & D is reported in ALATNET Newsletter part.
1. In Austria
First experiments with the NH version of ALADIN were carried out at ZAMG in cooperation with the University of Vienna. The work focused on high resolution simulations (<5km mesh size) in alpine areas. The NH model was run at different horizontal and vertical resolutions both on recent as well as historical cases such as, for example, from MAP-IOP 1997. (For further details contact wang++at++zamg.ac.at or bodo.ahrens++at++univie.ac.at).
As part of a simple statistical optimization of point forecasts of temperature, verification was done for the spatially interpolated forecast of temperature in Vienna, comparing direct model output (DMO) to observations of the synop station Vienna/Hohe Warte (11035). As an example, the distribution of these temperature differences for October 2000 is shown (bar graph). Overlayed is the fitted normal distribution (line graph) which gives a bias of -0.9°C and a standard deviation of 1.9°C. Based on analyses like the one shown, month-specific temperature corrections will be applied to the data before it is passed on to customers. (For further details contact stefan.greilberger++at++zamg.ac.ac).
2. In Belgium
Verification
Development of a systematic verification program of the Aladin Belgium forecasts. This includes the writing of a software package that can now be used for the calculation of the commonly used scores. This package keeps a database updated with data from Aladin and synoptic observations and soundings. It also contains a number of programs to calculate scores of the forecasts compared to the observations of the Belgian synoptic stations as well as the soundings in Uccle. It also allows producing some graphs such as scatter plots, error distribution bar charts and daily fluctuations of the scores. This package will be used to start the sending of monthly scores to the other contact points of the Aladin verification project.
3. In Bulgaria 
4. In Croatia
5. In Czech Republic 
Developments in physics
The package CYCORA bis was implemented first within the research library AL12_op5. Some preliminary tests were launched, namely on the case of 1997 flood. Compared to the 1997 operational run and to the recently recomputed CYCORA run one can say that all the three forecasts were of a very good quality. Still, there are a few interesting differences in the spatial distribution of precipitation patterns. More details can be asked to: Dagmar Merkova.
A search for a more realistic modeling of the PBL height continued, yet not offering a robust solution. More details can be asked to: Martina Tudor and Jean-Francois Geleyn.
This subject was tackled first by Mark Zagar and after his departure for the ”Far North”, Stefan Greilberger took over. The basic code was completed and tested, however, there is apparently still a coding bug. Therefore we cannot say at this stage whether the method is going to work or not. There are intermediate reports available in Prague. More details can be asked to: Stefan Greilberger, Thomas Haiden, Jean-François Geleyn.
Developments in the dynamics
A master thesis on ”Study of the applicability of Semi-Lagrangian schemes in NWP models with respect to the dependency on the increase of spatial resolution of the computational grid” has started. More details can be asked to: Alena Trojakova, Radmila Brozkova.
Developments in the verification
The tool ”VERAL” has been updated in order to compute the statistics with respect to the EWGLAM recommendations and the station list. More details can be asked to: Zuzana Huthova, Martin Janousek.
Technical developments
A first version of the graphical user interface to CHAGAL visualization program has been developed. The tool based on Tcl package (GNU) is designed to run on a Linux workstation and serve as a visual analyzer of ALADIN results, somehow more comfortable than the origina namelist-driven CHAGAL. After the tool reaches stable version it will be available also to other ALADIN sites. More details can be asked to: Tomas Kalibera and Martin Janousek.
6. In Hungary
The main developments were the sensitivity test using new local soil texture data over Hungary and further test of this aspect of the ISBA scheme using the 1D model version of ALADIN.
7. In Moldavia
8. In Morocco
The research and developpement work consisted mainly in :
9. In Poland 
10. Portugal
11. In Romania 
Comparison of the ALADIN results using different orography data sets (Steluta Alexandru)
The actual operational climatic files for ALADIN/Romania are obtained using GLOB95 orography data set and the Jerczynski function (option LNEWORO2=.T.) for the spectral fit. New climatic files have been built on GTOPO30 data set basis, with the same spectral fit (see also the February- March 2000 Toulouse stay report of Steluta Alexandru).
A comparison of the Aladin integration using the two climatic files has been carried out. Mainly we compared 10 m wind, temperature and humidity (at surface, 925, 850, 700 hPa), vertical velocity and precipitations. The first results have shown small differences for the above fields and for the objective verification scores as well (i.e. maximum differences for MSL pressure about 0.1 hPa, for 2m temperature about 2°). For temperature and humidity there is no preference in the maximum difference location; for wind and vertical velocities the higher differences have been observed in the south-eastern part of Romania and the western basin of the Black Sea.
The tests have been realised only for few situations (mainly for a stable stratification) and further experiments are necessary to choose between the two climatic data sets.
Tuning of the CYCORA-bis parameters (Doina Banciu)
The 1D model (for TOGA-COARE squall line of 22 February 1993) has been used for testing all changes of the CYCORA-BIS package and for the tuning of the free parameters of the deep convection parameterization scheme. Also some simulations have been carried out for the storm of 26 of December 1999 (with ARPEGE model) and CLEOPATRA case (with ALADIN model at 7.3-km resolution).
Preparation of an interface ALADIN-INPUFF pollutant dispersion model (Mihaela Caian)
Trajectory computation (Mihaela Caian)
The trajectories are calculated from simulated wind fields, with both horizontal and vertical components, derived from Aladin model. They can be calculated forward and backward in time:
The procedure for trajectory computation from the Aladin output files is ready and has been tested for 3hours interpolation time frequency for different option on vertical levels (pressure or height levels). For the tests, the averaging time, a free parameter of the procedure, was set to two-time-steps. The procedure output is available on the plane or on the sphere.
Tests on local topography impact (Mihaela Caian, Steluta Alexandru, Raluca Radu)
Two nested domains included into the operational domain for Aladin Romania have been used:
- D1: 22.8-25.2 E, 47.2-48.0 N
- D2: 23.38-23.75 E, 47.63-47.75N
Two parallel climatic files, using the GTOPO30 data set and a GIS data basis (with 500 and 200m resolution) have been created for each domain. The results of the Aladin model using the different climatic files have been compared for two types of integrations:
Hydrostatic for D1 + non hydrostatic for D2
Non hydrostatic for D1 + non hydrostatic for D2
Time interpolation effect on flux computation (Mihaela Caian, Anca Barbu)
It was been shown that for some cases the turbulent vertical exchange is strongly dependent on the forecast ranges chosen for interpolation. A smoothing operator based on the total flux amount normalization function is under test for the Media model input.
Improvement of coast representation (Steluta Alexandru)
For the tests concerning the improvement of the sea coast representation two new domains with different resolution have been created: one covering the whole Black Sea and another covering the half West side of the sea. Some sensitivity tests have been realized in order to study the impact of the variation of the parameters involved in the cost function. Also some new formulations of the Jerczynski cost function to introduce the sea fraction in the weight expression have been tried.
12.
In Slovakia
ALADIN to RODOS interface
Slovak Agency for Nuclear Safety (SANS) asked SHMI to develop interface which would enable to run RODOS model on ALADIN data. The interface was developed during year 2000, but RODOS model has not been tested on ALADIN data yet (technical problems on side of SANS).
RODOS system comprises 2 models:
These models use different types of input data:
Ascii data are obtained from outputs of ALADIN-SLOVAKIA. Fullpos is performed first. Simple data manipulations follow. FA-routines are used to read PF file.
GRIBs are transformed in 3 steps. In Prague, selected fields from ALADIN-LACE are converted to lower resolution and gribbed. Results are transferred to ZAMG. After downloading these GRIBs, transformation of geometry using GRIBEX package is performed. Final manipulations are done using wgrib/gribw utility.
Source code of ALADIN to RODOS interface will be released for other meteorological services. Export version is in preparation, it should be ready till the end of January 2001. However, since the interface was not written professionally, potential users will have to adapt it for their environment. It concerns mainly the makefile options and paths. Sources are in f90, C, and perl (version 5.003 or higher).
CANARI
There is an ongoing work on CANARI analysis, both in the frame of the diploma thesis and a preparatory work for future use of diagpack. The necessary code was ported on DEC ws: AL12 version is used for CANARI, and MA06 version for Mandalay.
Also the software for the conversion of the SYNOP data from the local netcdf database to the format required by Mandalay was written and tested. The surface CANARI analysis in ALADIN/SLOK and its impact on the forecast will be tested in the begining of 2001.
13. In Slovenia 
14.
Deported work by Météo-France people
 |
Home |