Abstract. This work exploits the potentiality of hail warning, based on single-polarization X-band weather radar measurements and tested on a large and well-documented data set of thunderstorm events in southern Italy near Naples. Even though X-band radars may suffer of two-way path attenuation especially at long ranges, due to their relatively low cost their use is rapidly increasing for short-range applications such as urban environments. To identify hail through radar measurements, two different methodologies have been selected and adapted to X-band data within the study area: one uses the Waldvogel (WAL) approach, whereas the other one uses the Vertically-Integrated Liquid Density (VIL-Density) product. The study aims at developing a Probability-of-Hail (POH) index in order to support hail risk management at urban scales. In order to find the optimal threshold values to discriminate between hail and severe rain, an extensive intercomparison between outcomes of the two methodologies and ground truth observations of hail has been performed, using a 2 x 2 contingency table and statistical scores.

The results show that both methods are accurate for hail detection in the area of interest, although VIL-Density product is less satisfactory than WAL method in terms of false alarm ratio. The relationship between the output of these two methodologies and POH has been derived through a heuristic approach, using a third-order polynomial fitting curve. As an example, the POH indexes have been applied for the thunderstorm event occurred on 21 July 2014, proving to be reliable for hail core detection.