|AROSA – Assimilation of Radio Occultation from Commercial Satellites over Austria|
|Project team:||Marc Schwärz (Senior Scientist)|
|Zentralanstalt für Meteorologie und Geodynamik (ZAMG), Vienna, AT; Lead|
|FFG-ALR (Austrian Research Promotion Agency - Austrian Aeronautics and Space Agency); ASAP-13 (Austrian Space Applications Programme 13th Call)|
May 2017 – Apr. 2019
Austria depends significantly on high quality, highly resolved weather forecasts, especially due to its complex orography, manifold landscapes and special meteorologically induced natural hazards in the alpine area and its important economic branches agriculture and tourism, which are strongly impacted by weather. The success of these forecasts is determined by a precise definition of the current state of the 3D atmosphere with highly resolved measurements due to the nonlinear nature of atmospheric processes.
Radio occultation (RO) methods exploit the bending of a radio signals on their way through the atmosphere by measuring the Doppler shift between a global navigation satellite system (GNSS) and a low earth orbit satellite (LEO) and their precise positions. The refraction of the signals enables atmospheric bending angle measurements that carry accurate information about temperature and moisture at high vertical resolution of a few hundred meters, including at stratospheric altitudes where conventional observations (aircraft and radio soundings) are relatively scarce. Such RO data now become available at increasing spatial density from commercially launched and maintained satellites of the company Spire Inc.
Within the scope of this exploratory project it was hence the aim to utilize the new occultation measurements of Spire Inc. to assimilate them for the first time into the high-resolution numerical weather prediction system AROME operated by ZAMG over Austria. So far, less dense RO data have been used for initializing some global weather forecast models (such as Arpège, GME, ECMWF-IFS), but for the high-resolution case of this project the adequacy of the observation operator simulating the measured data is rather crucial for success. Since the Spire Inc. RO data then became available at sufficient spatial density somewhat too late for this exploratory project’s timeframe, the analyses were carried out with comparable data of similar quality, in particular from the RO satellite constellation mission COSMIC.
The analyzed RO data were first pre-processed and evaluated for the model use (derivation of the “bending angles” and data quality assessment through validation) and then a “bending angle” error model was derived (observation error covariance matrix). For the data assimilation initially a 2D bending angle observation operator was available in AROME, which was developed for coarser resolutions. Based on case studies and a test period, the possible impact of the new observations on the analysis performance was studied and the potential of an operational application of the data within the AROME system over the Alpine region and Austria assessed. Through thinning of the other (non-RO) observations also the evaluation of the impact on similar forecasts in data poor regions outside Europe was possible.