Year of selection 2017
It is not just natural disasters that pose a risk to society and the economy. As a matter of fact, small scale weather events, such as severe rainfall and periods of snowmelt, are responsible for the majority (60%) of global insured losses. Anticipating such episodes and the chain of events that leads to adverse impacts is a challenging endeavour. « Current statistical models exhibit substantial limitations », Dr. Christian Rohrbeck points out when explaining the objective behind his postdoctoral project. « They fail to explain the weather dynamics leading to property losses ». The researcher thus aims to develop a new methodology that will better evaluate the risks induced by such weather events on a much more detailed regional level. In the wake of climate change, his goal is also to assess the impact of global warming on the frequency and severity of these risks. The outputs of this project will help society and the economy in several aspects, for instance, with a view to designing future protection measures.
« My proposed research project focuses on property damages. Specifically, I will analyse
insurance claims related to water-related weather events, like rainfall and snow-melt », Dr. Rohrbeck explains. « Current statistical models exhibit substantial limitations and fail to explain the weather dynamics leading to large monetary losses. The modelling is challenging due to complex nature and regional differences in the risk process, where the latter are related to geographical variations in climate, topology and socio-economic factors ». To illustrate this, he explains that « a certain amount of rain might be dangerous for one city and safe for another ». « The models thus need to account for spatial variations. In other words, they need to be focused, and deal with a wide range of statistics at the same time, », he continues. The flexible and rigorous, but computationally efficient, alternative statistical methods the aims to develop will help achieve this.
Better risk models for small-scale weather events
Dr. Christian Rohrbeck will develop his new methodology by building on the research conducted during his PhD and on two of his scientific papers from last year. He will extend the approaches produced then and create new ones, in order to successively reach the following aims : model the association between the number of claims and weather-metrics, between claim sizes and weather-metrics and assess the impact of climate change on property insurance claims. The performance of these models will be assessed based on an insurance data set which covers all Norwegian municipalities.
Both small and large weather related insurance costs are rising due to global warming and climate change. « Large aspects of society and the economy are weather-sensitive, and insurances against undesirable weather events are an important economical factor », Dr. Christian Rohrbeck stresses. In this sense, his project will have a important impact on the scientific community, as well as in the area of public debate. Not only will his output profit the insurance industry and stimulate new research, it will also inform politicians and decision-makers about how they should prepare for the future.