Quality of precipitation prediction by the NWP model WRF-ARW with preliminary data assimilation

• Published in: IOP conference series. Earth and environmental science Vol. 211; no. 1; pp. 12061 - 12069
• Main Authors: Kostarev, S V, Vetrov, A L
• Format: Journal Article
• Language: English
• Published: IOP Publishing 17.12.2018
• ISSN: 1755-1307; 1755-1315; 1755-1315
• DOI: 10.1088/1755-1315/211/1/012061

This paper represents a new method of improving a precipitation prediction by the WRF-ARW model which is based on a preliminary assimilation of GFS objective analysis and forecast data. The article is devoted to a comparison of the quality of precipitation prediction by the WRF-ARW run in 2 modes: using a preliminary data assimilation and using a common approach. It was found out that the preliminary assimilation of GFS objective analysis and forecast data allows one to improve the prediction quality of precipitation fact, which is assessed by precipitation fact (absence), forecast reliability and precipitation fact (absence) warning. These quality characteristics increase by 2-4% in case of preliminary data assimilation. An increase in the prediction quality of precipitation amounts is observed using preliminary data assimilation. The absolute error mean of precipitation amounts forecast is 2.13 and 2.03 mm using the preliminary data assimilation and the standard approach, respectively. Furthermore, the preliminary data assimilation helps improve the prediction quality of heavy precipitation (≥ 15 mm/12 h) fact. The heavy precipitation forecast reliability and warning increase by 5 and 9%, respectively, using the preliminary data assimilation. Additional characteristics of heavy precipitation prediction quality, i.e. Pearcy-Obukhov and ETS criteria, increase by 0.10 and 0.03 in comparison with the standard approach, respectively.