A ten-year statistical radar analysis of an operational hail suppression program in Alberta

Hailstorms represent a significant natural hazard worldwide and can result in severe socioeconomic impacts. In Canada, the province of Alberta experiences the highest rate of hail occurrences causing tremendous damage to cities, properties, vehicles and crops. Cloud seeding is often adopted to reduce the hailstorm impacts; however, the efficacy of this mitigation measure is still unclear. In this study, a ten-year set of hailstorm events (2011-2020) in Alberta was analyzed using radar-based records to evaluate the potential effects of seeding. The radar completed a full volume-scan every four minutes, and a three-dimensional radar reflectivity thresholdbased method was utilized to identify hailstorms. A single 3D storm is tracked over time and the ensemble is called a "storm track" in this contribution. 176 storm tracks persisting for more than one hour were considered for further analysis. We used two radar-derived metrics as proxies for hail damage potential, namely, VILmax (Vertically Integrated Liquid calculated from the maximum reflectivity profile in the storm) and hail mean coverage area (defined as the mean area with reflectivity greater than 60 dBZ). Seeding was performed by releasing silver iodide aerosols from aircraft, and seeded scans of the storm were identified based on the location and seeding time. The efficacy of seeding was evaluated by comparing the characteristics of the storms before, during and after seeding. Results show that for nearly 60% of cases, the median values of the largest (quantile 99) VILmax and mean area corresponding to the seeded scans are lower than those of the unseeded portions. For around 8% and 20% of cases, there is no change, while for approximately 17% to 30% of cases, the values are higher. The effects of seeding are more pronounced after 30 min than at the initiation of seeding. Furthermore, statistical analyses using Mann-Whitney and Wilcoxon's tests reveal that differences in VILmax and mean area between the seeded and unseeded scan groups are statistically significant. These findings suggest that cloud seeding could potentially influence the hail damage potential of storms, particularly those with high VILmax values.