Assessment of the impact of key terrestrial observing systems using DMI‐HIRLAM

Following a EUCOS decision to make an assessment of the impact on NWP forecasts of different components of the current observing systems in various combinations, a number of OSEs (Observing System Experiments) have been made by a number of NWP centres, with some running global models and some limited‐area models. The Danish Meteorological Institute (DMI) carried out OSEs with the limited‐area model HIRLAM. The following runs were made by DMI (two periods, one month each): (1) A baseline system (BL); (2) BL + all aircraft wind and temperature data; (3) BL + non‐GUAN radiosonde winds; (4) BL + non‐GUAN radiosonde temperature and wind; (5) BL + wind profiler; (6) as (4) + aircraft wind and temperature; (7) as (4) + non‐GUAN radiosonde humidity; (8) as BL + all in situ data (full combined system); and (9) BL + E‐AMDAR only (no ACARS and no traditional AIREPs). The lateral boundaries for these OSEs were provided by runs made by ECMWF, which was one of the centres to make OSEs with a global model. The main conclusions are that the radiosonde data are the most important data, closely followed by the aircraft data, and that aircraft data and radiosonde data are complementary and not redundant data. Furthermore the results show that it is important to have both wind and temperature profile data; wind data alone produce much poorer impact. Copyright © 2008 Royal Meteorological Society