Methodology to produce super-observations from radar wind raw data
Weather radars are often used for the remote collection of data. A Doppler radar uses electromagnetic waves to investigate the atmospheric properties by transmitting electromagnetic wave pulses and receiving reflected echoes. Radars can sense data with high spatial and temporal resolution, improving the quality of high resolution limited area weather forecast. However radar data, despite its great potential due to its spherical geometry and high measurement density, has a limited use. This is mainly due to certain limitations related to the maximum velocity of scattering particles which can be resolved and the maximum radar range. Since the radar beam broadens with increasing range, an atmospheric phenomenon could be over-sampled close to the radar and under-sampled in other parts of the domain, creating scale discrepancies. In order to overcome these problems, nine European institutes have developed a processing procedure for raw radar radial wind data. The aim of the EU-project CARPE DIEM is to assess the potential gain from incorporating radar data in the NWP process. The procedure includes averaging raw data in polar space, named as super-observations, and a suitable filter of super-observations to better match the time scales of the model. In addition, a four-dimensional observation operator has been developed for assimilation in NWP, the application of which, when utilising radial wind data, has experimentally indicated some gain in forecast quality. Precipitation and wind forecasts can be significantly improved by assimilating radar observations in NWP. Case studies using radar measurements show promising results, encouraging extended studies for their confirmation and further development.
