When a hail event lasting just minutes hit southern Germany on July 28, 2013, it generated a US$3.9 billion insured loss. Some 80,000 buildings and tens of thousands of automobiles were damaged — many severely. Such a high hail loss was unprecedented in Europe. It was this event that represented the wake-up call for the insurance and reinsurance industry to think harder about the risk due to severe convective storms.
I had the privilege to lead my team of seven modelers from RMS as we attended the ninth European Conference on Severe Storms (ECSS) which took place between September 18-22 this year in Pula, Croatia. This biannual conference sees hundreds of scientists predominantly from Europe and the U.S. but also from other parts of the world. Hail was prominent on the event agenda among the scientific community, and with rising interest in hail damage there is now also a considerable participation from the reinsurance industry. Munich Re is the main sponsor of the conference.
Delegates at the European Conference on Severe Storms 2017. Image credit: ESSL – Thomas Schreiner
The event was very interesting for RMS, as we always want to build models based on the best scientific knowledge, using the most up-to-date methods and data whenever possible. To take the most effective approach to model risk, we also recognize that methods must be tailored or developed. And in these cases, we can successfully contribute to the scientific discussion.
New Insight on Hail
Our team contributed five out of 186 presentations across the conference. All our contributions raised considerable interest among the experts of the various fields. Our hail-tracking algorithm — based on graph theory, was discussed for hours as an alternative to other well-established methods for the identification of hail using radar reflectivity, because of its simple applicability and especially its robust results.
One very difficult subject to tackle is the intensity distribution and footprint size of convective gusts. Our detailed model runs and calibration with station observations is novel and was highly appreciated by the experts working in this field.
With respect to hail there are several ways to express intensity. Radar reflectivity, maximum hailstone size and hail-kinetic energy are some of them. In one of our contributions we showed how they are linked to each other. Based on the hailstone size spectrum and observational data from hail pads, we can estimate the probability that a subject at risk gets hit with a certain number of hailstones of a certain size range given the maximum hailstone size. This gives concrete meaning to a maximum hailstone size and is a rigorous step for the engineering development of hail vulnerability curves.
Clients and the scientific community appreciated our contributions and wanted to know more details which we discussed in many one-to-one meetings, during poster sessions, coffee breaks and the conference dinner.
We always welcome the opportunity to openly present the scientific details of our approach with the community, to get feedback and criticism. As an organization, this is a deliberate strategy on our part, we know how important discussion and debate is. We understand that sometimes it can be frustrating for the scientific community to watch a high-level presentation where the details of the underlying strategy and methods are not revealed.
After having left the academic world some 17 years ago, it was great to see how my team contributed to so many different aspects of severe convective storms. And even though we are working for and dedicated to the insurance industry, we are still part of the scientific community and we look forward to continued lively discussion, debate, and knowledge sharing at events of this kind in the future.