Stefano Zanardo, Principal Modeler, RMS
Ludovico Nicotina, Senior Director – Modeling, RMS
Arno Hilberts, Vice President, Model Development, RMS
Steve Jewson, Scientific Research Consultant, RMS
The North Atlantic Oscillation (NAO) describes the fluctuations in the difference of atmospheric pressure at sea level between two semi-permanent centers of low and high pressure in the North Atlantic: the Icelandic Low and the Azores High. Fluctuations between these centers control the strength and direction of westerly winds and location of storm tracks across the North Atlantic.
Why is this important? The NAO signal is Europe’s dominant mode of climate variability and correlates highly with European precipitation patterns. Typically, when the NAO is positive – characterized by a higher than average pressure difference between low and high latitudes of the Northern Hemisphere, Northern Europe experiences strong westerly winds. This causes stormier and wetter than usual conditions in Northern Europe, while Southern Europe is drier and colder than usual.
In contrast, when the NAO is negative, Southern Europe experiences westerly winds and the meteorological pattern is somewhat opposite, with Southern Europe being generally wetter than average. The NAO is significantly stronger in winter than in the other seasons, therefore, most studies on the NAO focus on winter months, when the influence of the NAO on surface temperature and precipitation is highest.
When climate patterns result in changing prevailing conditions, such as increased storm activity and rainfall, it is important to understand their effect in relation to the severity of flood events – responsible for significant property damage, business disruption and loss of life in Europe. And there is a need to understand its ongoing impact as the climate and the distribution of exposures change over time.
Understanding NAO and its Impact on Europe Flood Losses
Although the structure of flood losses is arguably related to large scale climatic patterns, this inter-connection is not always well understood. While it is well known that large scale climatic patterns control meteorological events, it is not always clear whether this connection can be extended to the occurrence of flood events and the associated losses.
So, despite well-established knowledge of the effect of NAO on precipitation patterns, hypotheses on relationships between the NAO and catastrophic flood events and losses have not been considered in scientific literature. This can be attributed to the lack of data on historical flood events but also to the fact that appropriate, large scale loss models are typically not available to the research community.
In a new study, published in Geophysical Research Letters and recently featured in Nature Research Highlights, we have now shown that there is indeed a significant relationship between the NAO and flood losses.
The analysis suggests that there is a clear relationship between the occurrence of catastrophic flood events across Europe and the NAO signal. In Northern Europe, the majority of historic winter floods occurred during a positive NAO state, whereas the majority of summer floods occurred during a negative NAO state. Critically, we observe that the average flood loss during opposite NAO states can differ by up to 50 percent.
These results are not solely justified by the fact that the NAO modulates precipitation amounts and, in turn, precipitation modulates flood events. Indeed, the same precipitation event may or may not generate a flood, depending on soil saturation and therefore on prior precipitation. The NAO exhibits a certain level of persistency and there can be prolonged periods with a predominant NAO state, associated with long periods of wet or dry conditions, which increases its impact on flood generation. As an example, persistent positive NAO during the 2015/2016 winter contributed to wetter than usual conditions in Northern Europe as an unusual cluster of storms hit the U.K. and Ireland. This caused catastrophic floods throughout the region and, consequently, significant damage and disruption.
Our study used the RMS Europe Inland Flood HD models. Based on latest scientific understanding of flood processes and advanced modeling techniques, the RMS model generates a large set of potential extreme events and quantifies the associated damages. Thousands of years of simulated precipitation is used to drive a modeling cascade that describes rainfall-runoff, river routing and inundation processes, including a suitable model for flood defenses along major rivers and their failure probabilities.
Our findings on the impact of the NAO on financial flood losses have a number of implications. From a short-term point of view, knowledge of these relationships can improve financial preparedness and capital allocation for disaster funds. Knowing that in a certain region, positive/negative NAO states are likely to be connected to higher/lower flood losses, allows stakeholders such as insurance companies and governments to budget financial resources accordingly, if skilful NAO forecasts are available.
Therefore, the earlier we can predict lower or higher NAO periods the better the preparedness will be. Based on the present study, NAO predictions will be instrumental towards improving preparedness and resilience to catastrophic floods.
From a longer-term point of view, the spatial patterns observed in the loss-NAO relationship indicate that the NAO states constitute an additional, undetected source for the well documented spatial correlation of flood risk. Since insurance schemes and public recovery funds rely on a deep understanding of loss correlation structures, this effect should be accounted for in future projections of flood risk.
Moreover, these results are particularly relevant in the context of climate change. Past NAO circulation patterns varied as part of changes in Northern Hemisphere climate and future changes can therefore be expected to have an impact on the NAO pattern. Our study shows that such changes would in turn affect economic losses due to floods, and this effect can be positive or negative depending on the season and region. These results can inform financial preparedness and disaster fund allocation, as stakeholders can distribute resources more effectively.