Dual polarization radar Lagrangian parameters: a statistics-based probabilistic nowcasting model

• Published in: Natural hazards (Dordrecht) 2017-11, Vol.89 (2), p.705-721
• Main Authors: Medina, Bruno Lisbôa, Machado, Luiz A. T.
• Format: Article
• Language: English
• Subjects: Civil Engineering; Correlation coefficient; Correlation coefficients; Crystals; Dual polarization radar; Earth and Environmental Science; Earth Sciences; Echoes; Environmental Management; Geophysics/Geodesy; Geotechnical Engineering & Applied Earth Sciences; Graupel; Growth; Hydrogeology; Hydrometeors; Ice; Ice crystals; Mathematical models; Meteorology; Natural Hazards; Nowcasting; Original Paper; Parameters; Physical characteristics; Physical properties; Polarimetric radar; Polarization; Radar; Radar echoes; Radar polarimetry; Rain; Rainfall; Reflectance; Statistical methods; Statistics; Storm development; Storms; Supercooled water; Weather forecasting
• ISSN: 0921-030X; 1573-0840
• DOI: 10.1007/s11069-017-2988-y

The aim of this study is to present a statistics-based Lagrangian nowcasting model to predict intense rainfall convective events based on dual polarization radar parameters. The data employed in this study are from X-band radar collected during the CHUVA-Vale campaign from November 2011 to March 2012 in southeast Brazil. The model was designed to catch the important physical characteristics of storms, such as the presence of supercooled water above 0 °C isotherm, vertical ice crystals in high levels, graupel development in the mixed-phase layer and storm vertical growth, using polarimetric radar in the mixed-phase layer. These parameters are based on different polarimetric radar quantities in the mixed phase, such as negative differential reflectivity ( Z DR ) and specific differential phase ( K DP ), low correlation coefficient ( ρ hv ) and high reflectivity Z h values. Storms were tracked to allow the Lagrangian temporal derivation. The model is based on the estimation of the proportion of radar echo volume in the mixed phase that is likely to be associated with intense storm hydrometeors. Thirteen parameters are used in this probabilistic nowcasting model, which is able to predict the potential for future storm development. The model distinguishes two different categories of storms, intense and non-intense rain cell events by determining how many parameters reach the “intense” storm threshold.