Proceedings - C-SRNWP Project

Data Assimilation for LM
Method: Nudging towards observations
Implementation: Continuous cycle
Analyzed variables: horizontal wind vector, potential temperature, relative humidity, 'nearsurface'
pressure (i.e. at the lowest model level)
Observations:
SYNOP, SHIP,DRIBU: station pressure, wind (stations below 100m above msl) and humidity
TEMP,PILOT: wind,temperature: all standard levels up to 300 hPa; humidity:all levels up to
300 hPa; geopotential used for one “near-surface” pressure increment.
AIRCRAFT: all wind and temperature data
WIND PROFILER: all wind data (not included in blacklisted stations)
Quality Control: Comparison with the model fields from the assimilation run itself.
A Latent Heat Nudging scheme is also implemented for testing.
3. Recent developments.
During this last year, five new version of LM have been released. The present one is the
version 3.21 which is going to became the future LM reference version 4.1. The most
relevant changes with respect to last year are:
– The implementation of snow density to the multilayer soil model
– Some changes and tuning in the assimilation scheme. Much work has been
done to allow the assimilation of temperature and humidity profiles retrieved
from satellite data by means of a 1-DVAR technique (1-DVAR Special
Project). This scheme is now under tuning and extensive testing.
– The implementation of the FLake Model
– Some changes to Runge-Kutta Scheme (RK Priority Project)
– The complete harmonization with the LMK code
– The adaptation of the code and of the input files to support the run of LM in
Ensemble Mode by parameters perturbations (SREPS Special Project)
Another important and relevant upgrading of LM was the Climate Mode extension of the code
to allow and support the use of the model in climate applications. Major changes were related
to the restart of the model and to the output now available also in NetCDF format. This
cooperation with the CLM group already gave some useful feedbacks to the COSMO
community highlighting some technical inconsistencies and bugs of the model and testing the
long-term behaviour of LM. For more information see the presentation by Andrea Will
available at the 2006 COSMO general Meeting WEB site.
4. LM operational applications.
The LM runs operationally at the different COSMO Countries. The operational integration
domains, the correspondent horizontal resolution and the driving models are reported in figure
1.
For more details please see the national reports in this Newsletter issue.
35