Natural hazard insurance outcomes at national, regional and local scales: A comparison between Sweden and Portugal.

This study addresses the role of natural hazard insurance in two European countries with different insurance markets and socioeconomic conditions: Sweden and Portugal. The analyses were conducted at the national, regional (Southern Sweden and Lisbon Metropolitan Area - LMA), and local (Malmö and Lisbon cities) scales. Most damage caused by weather and climate-related (WCR) hazards during the 1980-2019 period was not covered by insurance companies in Sweden (71%) and Portugal (91%). An insurance affordability analysis was performed using income for the national and regional scales. Unaffordability is higher in Southern Sweden than in LMA, implying that better socioeconomic conditions do not necessarily mean a higher average capacity to pay for insurance. At the local scale, urban flooding was analysed for Malmö (1996-2019) and Lisbon (2000-2011) using insurance databases, in which the most relevant 21st century rainfall events for each city are included (2014 and 2008, respectively). The influence of terrain features on flooding claims and payouts was determined using Geographic Information Systems (GIS) spatial analyses. The flat Malmö favours ponding and extensive flooding, while the distance to the drainage network and flow accumulation are key factors to promote flooding along valley bottoms in the hilly Lisbon. Flooding hotspots tend to result from a combination of higher depths/lower velocities (accumulation of floodwaters and ponding) and not from a pattern of lower depths/higher velocities (shallow overland flow). More detailed data on insurance, flooding, and socioeconomic conditions, at regional and mainly local scales, is needed to improve affordability and urban flooding risk assessments.

Flood insurance is a driver of population growth in European floodplains.

Future flood risk assessments typically focus on changing hazard conditions as a result of climate change, where flood exposure is assumed to remain static or develop according to exogenous scenarios. However, this study presents a method to project future riverine flood risk in Europe by simulating population growth in floodplains, where households' settlement location decisions endogenously depend on environmental and institutional factors, including amenities associated with river proximity, riverine flood risk, and insurance against this risk. Our results show that population growth in European floodplains and, consequently, rising riverine flood risk are considerably higher when the dis-amenity caused by flood risk is offset by insurance. This outcome is particularly evident in countries where flood risk is covered collectively and notably less where premiums reflect the risk of individual households.