9. Operational version in Hungary

(more details horanyi++at++met.hu)

During the second part of 2002 we have successfully installed a new ALADIN domain together with the switch to a direct coupling to the ARPEGE model as lateral boundary conditions (see details in the previous Newsletter). The main activities during the first half of 2003 were devoted to the further experimentation of the three-dimensional variational data assimilation scheme (3d-var) for the ALADIN model.

The most important operational activities around the local version of the ALADIN model in Budapest are detailed hereafter:

  • -- The operational scripts were renewed with the help of a professional software developer team. The new scripts became very modular and systematic (however their readability had been significantly decreased).
  • -- The RETIM satellite data transfer was put into use at the end of April and systematic comparison of the data transfer through RETIM and internet-ftp was carried out afterwards. The results showed that there is a higher reliability found in the RETIM dissemination, however the speed of the transfer is a bit slower than the speed of the internet capacity (detailed statistics can be provided on request for the more interested readers). Based on these measurements we have decided that we do use operationally the initial and lateral boundary files provided by RETIM (since mid-June). The internet solution was kept as a backup.
  • -- The 3d-var data assimilation cycle was continuously running in parallel suite (with 48 forecast once per day) and the results of the double suite was systematically compared to the reference solution (dynamical adaptation). The main outcome of the comparison is briefly described in the "ALADIN developments" chapter of the same Newsletter. The classical double suite used only SYNOP and TEMP observations, nevertheless some further studies were carried out using additional ATOVS and AMDAR data in the process of 3d-var.
  • -- During the spring we started to operationally monitor the performance of the ALADIN operational suite. It means that beside the graphical tool available for the operators (about the successful completion of the model run), the developers have to their disposal a laptop and a mobile phone, and a remote-login is ensured in order to make intervention in the operational suite in case of necessity.
  • -- The loadleveler job scheduler was adjusted several times, nevertheless its functionality is far from being optimal, preventing the full usage of all the available resources of our IBM Regatta computer.
  • -- We have introduced just very few new elements to the operational model suite: we started to make verification over the ARPEGE LBC files (for direct comparison to the ALADIN/HU version) and we have created special post-processing for the RODOS environmental application and for another dispersion model predicting low level pollution.
  • -- AL25 model version was installed and partly validated, but not introduced operationally.