The limitations of observational data available for the study of damaging weather conditions (e.g., storms and extreme temperature events) are discussed. Crop and property insurance loss records are advocated as a potential supplement to traditional weather observations, as they integrate specific information about the spatial dimension of damaging weather conditions and the cost of damage they cause. Insurance loss data may also be analyzed in combination with meteorological data sets to derive indicator variables for the detection of damaging weather events.
Two sets of insurance data are described. One record provides adjusted property losses associated with “catastrophic” weather events since 1949, and the other is an index of the amount of crop-hail losses per year since 1948. Additionally, an example of the benefits of the combination of insurance and meteorological data is presented through a selection of results from a recent study of freezing temperatures in the southeastern United States and associated insurance claims related to pipe bursting.
If insurance data are to be applied in the future in similar studies of damaging weather conditions, it is essential that the insurance industry continues to collect and adjust loss data and periodically confirm that adjustment factors are temporally consistent.