Seminar: Matic Šavli: Assimilation of horizontal line-of-sight winds with a mesoscale EnKF data assimilation system over the northern Atlantic and Europe

Thursday, 1st March 2018, 13.15
Room F5, 1st floor, FMF, Jadranska 19, Ljubljana

Meteorology Group will host a seminar presented by Matic Šavli. The seminar is open to the public, everyone is kindly invited. The seminar will be in English.

Abstract:

The forthcoming ADM-Aeolus mission of the European Space Agency will provide global coverage of vertical profiles of horizontal line-of-sight (HLOS) wind twice per day. In previous studies, the potential impact of HLOS winds has been extensively evaluated using the ECMWF variational data assimilation system and a significant impact was reported over the tropics. Over Europe, a minor impact was obtained for short forecast ranges but a positive impact was sustained in the medium range.
As part of the preparation for the use of HLOS winds in mesoscale NWP systems in Europe, we shall examine the impact of HLOS winds in an ensemble Kalman filter (EnKF) limited area data assimilation system over Europe and the North Atlantic. The study uses the Weather Research and Forecast model (WRF) and the ensemble adjustment Kalman filter (EAKF) from the Data Assimilation Research Testbed (DART). The Observing System Simulation Experiment (OSSE) framework was developed with a 50-member EAKF nested in the operational 50-member ensemble prediction system of ECMWF using model-level data available twice per day.
The importance of the flow-dependent background-error covariances was explored using experiments without covariance inflation. The value of HLOS winds is compared with a single wind component and full wind information. In areas of strong covariances such as along fronts in the Atlantic, multivariate information provides significant analysis increments, especially if the HLOS observation is aligned along the front. It is furthermore shown that the assimilation of wind information in terms of the HLOS component may produce better analyses than the assimilation of either the zonal or meridional winds.