Normalised losses from river flood disasters (CLIM 040) - Assessment published Sep 2008

Key policy question: .

Flood losses per thousand of GDP in the EU 1970-2005

Note: The figure shows flood losses per thousand of GDP in the EU 1970-2005

Data source:

Barredo, J. I., 2007. Major flood disasters in Europe: 19502005. Natural Hazards 42: 125148

Downloads and more info

Number of casualties caused by flood disasters in the EU 1970-2005

Note: The figure shows the number of casualties caused by flood disasters in the EU, 1970-2005

Data source:

Barredo, J. I., 2007. Major flood disasters in Europe: 19502005. Natural Hazards 42: 125148

Downloads and more info

Projected change in damage of river floods with a 100-year return period between 2071-2100 and 1961-1990

Note: Note: Model calculation using the IPCC SRES scenario A2 and NUTS2 level.

Data source:

JRC PESETA project. http://peseta.jrc.ec.europa.eu/docs/Riverfloods.html

Downloads and more info

Key assessment

Past trends

Flood disasters in Europe increased in number and amount of loss from the 1970s to the 2000s. The number of major flood disasters during the last 16 years (between 1990 and 2005) is more than twice that between 1970 and 1989 (Barredo, 2007). When assessing flood losses it is important to compensate for changes in asset values and exposure over time. Failure to adjust for economic factors results in loss amounts that are not directly comparable over time and a pronounced ever-increasing trend for purely economic reasons (Hoppe and Pielke Jr, 2006; Muir Wood et al., 2006). Figure 1 therefore shows the costs of flood losses in Europe as a percentage of GDP. A continuous increase is observed in the decadal average of flood damage expressed in this way.
In fact in the period 1970-1999 the trend in EU flood losses was not statistically significant, and the increase registered in the last sub-period is a consequence of one single event, the floods in central Europe in the summer of 2002. However, even though evidence indicates that the growth of flood losses in recent decades is related to both societal and climatic factors, the shares are unclear (Pielke Jr and Downton, 2000; Barredo, 2007). It is therefore still not possible to determine the proportion of the increase in damage that might be attributed either to climate change or to societal change and economic development (Hoppe and Pielke Jr, 2006). There is agreement that climate change cannot be regarded as the dominant factor for increasing flood losses. In addition there are no conclusive studies that confirm the hypothesis of changes in the occurrence of extreme river flows in Europe. In a hypothetical scenario without climate change, total flood losses will continue to increase as consequence of societal and economic factors such as increase in exposure and vulnerability (Pielke Jr and Downton, 2000). Figure  2 shows the yearly number of deaths resulting from floods in Europe for the period 1970-2005. There is no clear trend. The number of deaths is very dependent on single events, as for the events of 1970 in Romania and Hungary, 1973 in Spain, and 1998 in Italy. In recent decades, early warning systems and prevention measures have improved evacuation mechanisms in the many areas exposed to floods.
The issue of extreme precipitation and surface water flooding (heavy rainfall and insufficient capacity of drainage systems) is also worth further investigation since this is already causing problems while not being well enough understood in terms of risk mapping. It has been estimated that the 2007 summer floods in the United Kingdom were caused mainly by surface water flooding and inadequate drainage (roughly 60 % of the losses) while the rest was caused by river flooding.

Projections

Some preliminary estimates (ABI, 2005) indicate that annual flood losses in Europe could rise to EUR 100-120 billion (tenfold) by the end of the century under high emissions scenarios. Hall et al. (2005) presented a national-scale assessment for England and Wales, and predicted an up to 20-fold increase in losses by the 2080s in the scenario with the highest economic growth and no adaptation. These results include changes in sea-level rise, increasing precipitation and increasing economic vulnerability. More detailed disaggregated work under the PESETA project has modelled changes in river flows in a changing climate in Europe, studying two river catchments in detail.

• For the Upper Danube the estimated total damage of a 100-year flood is projected to increase by 2100 by around 40 % of the current damage estimate (an increase of EUR 18.5 billion) for the high emission scenario (A2) and around 19 % for the intermediate emission scenario (B2). The number of people affected is projected to increase by 242 000 (around 11 %) for the A2, and 135 000 (around 6 %) for the B2 scenario.
• For the Meuse, the potential damage of a 100-year flood is projected to increase by about 14 % for the A2 scenario and about 11 % for the B2 scenario.

These regional studies have been expanded for river flooding EU-wide. Figure 3 shows the percentage change in economic damage for floods with 100-year return period for the SRES A2 scenario.
A number of uncertainties in these river catchment and Europe-wide results should, however, be highlighted. First, the numbers are the combined effect of the climate and socio-economic effects, and second, they do not include existing or any future flood protection and management measures, so strictly speaking they are a measure of potential exposure, not impacts (though they may underestimate potential losses by not incorporating changes in exposure). This highlights a broad issue with climate and socio-economic analysis of future flood risks. Research into flood risks in the Netherlands indicates that potential economic losses from flooding (river and sea) as a result of socio-economic change could increase by 22-45 % in 2040 (WL Delft Hydraulics, 2007). The particular role of climate change was not taken into account, because of unknown effects on flood severity and frequency. Moreover, socio-economic factors are expected to dominate future loss records, and will continue to complicate normalisation studies, because of the large inaccuracies associated with actual loss estimates, compared with geophysical data on extreme weather itself (Pielke Jr, 2007).