Hailfall: the relationship between radar-derived hail kinetic energy and hail damage to buildings

Relationships between radar-derived hail kinetic energy and hail damage to residential and agricultural buildings are presented for nine hail cells that occurred over the Swiss Mittelland (1992–1999). Hail kinetic energy (EKINPIX) was calculated from C-band Doppler radar Constant Altitude Plan Position Indicators (CAPPIs) at a height of 1.5 km and daily hail damage claim data were available per Swiss community through several cantonal building insurance companies. In order to derive point-to-point relationships, hailfall and damage variables were triangulated on a 2×2-km grid and cross-correlation coefficients were then calculated for entire hail cells. sults show that high season hail storms (15 June–15 August) produced higher mean damage than low season storms (before and after) and weighted logistic damage functions produced correlation coefficients of 0.83 (0.86) for high (low) season storms over residential buildings and 0.79 (0.88) over agricultural buildings. The relationship between EKINPIX and total loss ratios (damage in relation to total sums insured) reveals no seasonal difference in hailstorm intensities and nonweighted logistic functions yield correlation coefficients of 0.85 (0.84) for residential (agricultural) buildings. The comparison between actual and predicted losses showed that nonweighted functions overestimated actual damage and that altered damage functions (flatter slope gradients) predicted damages that are in the range of the actual losses. Damage functions are calibrated using EKINPIX, total loss ratios and actual losses from two independent hail cells. One hail cell showed that damage induced through rainfall that followed hailfall multiplied the final loss considerably. The valuable agreement between radar-measured hailfall intensity and total loss ratios suggests that the established damage functions could be used by insurance companies to derive possible maximal hail losses to a building portfolio of interest.