The costs and impacts of the winter 2013 / 14 floods in England and Wales

Between December 2013 and March 2014, the UK witnessed heavy and prolonged rainfall, including the wettest January on record in parts of the country and around twice the average monthly rainfall in other locations. The East Coast was also affected by the largest coastal surge since 1953. These events resulted in significant coastal damage and prolonged fluvial and groundwater flooding affecting a large expanse of England and Wales. During this period many properties, crucial transport infrastructure and farmland were flooded. This study for the UK Environment Agency, led by Risk & Policy Analysts (RPA) in association with HaskoningDHV UK, John Chatterton Associates and Morris Resource Economics, reviewed more than 500 documents, articles and reports and requested information from more than 640 individuals and organisations. Through careful consideration of the reliability of each and every piece of data, we were able to calculate a best estimate for the economic costs of the winter 2013/14 floods (in 2014 values) of , with uncertainty resulting in a range of between .3 . Residential properties suffered the greatest proportion of damages, with 25% of total damages (best estimate of million incurred by 10,465 properties).


Introduction
From December 2013 to March 2014, England and Wales experienced a sustained period of flooding.This was caused by a continual and unusual period of winter storms.These storms resulted in significant coastal damage and prolonged flooding from fluvial, pluvial, coastal and groundwater sources.Key features of the floods included storm surges and waves causing coastal damage in both England and Wales, and persistent rainfall (particularly during December and January), which the MetOffice considers to be one of, if not the most, exceptional period of winter rainfall in at least 248 years [1].In total, there were more than 700 floods recorded in England and Wales, with 160 of these having a duration of 11 days or longer [2].Figure 1 shows the extent of flooding.This paper summarises the approach used to estimate the range of impacts and to calculate the financial and economic damages resulting from flooding in England and Wales during the winter of 2013/14.
The overall aim of the study was to identify how damages from flood events could be more robustly estimated.
As such, identification of sources and magnitude of uncertainty within the estimates was an integral part of the assessment process.Also important was the identification of the extent to which the magnitude and distribution of damages differed from those on previous flood events, such as in 2007 in England and Wales.This included consideration of damages from different sources of flooding: fluvial, pluvial, groundwater and coastal flooding, identifying who was impacted and to what extent.Table 3 shows that the best estimate for most categories is approximately the mid-point in the range.There are two categories where this is not the case though.For emergency services, the best estimate is based on actual damages incurred but there may be some data gaps (for example, costs to the military).However, the military recovered some of its costs from other authorities so these costs may be captured elsewhere.Since it was known that the best estimate may be an under-estimate, it was considered conservative and no further reduction in damages was made for the low end of the range.A similar approach was used for public health and welfare.4 and Figures 3 and 4 all show the large extent of flooding in the winter of 2013/14.There are also differences between the spread of residential properties and business that were affected.The proportion of businesses affected in the East of England (12%) was higher than the proportion of residential properties (6%).This reflects the impacts of the coastal surge, with there generally being a larger number of businesses along the coastal frontage.This is also reflected in the damages (see Figure 2) where 63% of damages to businesses occurred as a result of coastal, rather than fluvial flooding.

Comparison of the approach and results with the European commission guidelines
The approach used (as described above) was designed to be consistent with that applied following flooding in England in 2007 [4].This then allows for comparison between the two sets of damages.The approach can also be compared against other methods of damage estimation.For example, the European Commission LGHQWLILHV WKDW µsystematically collected, comparable and robust disaster damage and loss data are an essential element of the risk assessment and management SURFHVVHV ¶.To improve consistency, guidelines were prepared setting out a minimum set of loss indicators.

These include human loss indicators, damage indicators and economic loss indicators. The human loss indicators cover [5]:
-directly affected people: exposed people in terms of impacts on their livelihood immediately after the disaster (displaced, isolated, impaired) and impacts on their physical integrity (injuries).
-indirectly affected people: those who suffer indirect effects either within or outside the affected area.
-deaths: number of people who died during the disaster or sometime after as a direct result of the disaster -missing: number of people whose whereabouts since the disaster are unknown (presumed dead).

The damage indicators capture [5]:
-houses destroyed: number of household units levelled, buried, collapsed or damaged to the extent that they are no longer habitable/repairable. -houses damaged: number of household units with minor damage, that may continue to be lived in although they may require some repair or cleaning.
-education centres: the number of schools, kindergartens, colleges, universities, training centres, etc. destroyed or directly damaged.
-health facilities: the number of health centres, clinics, local and regional hospitals destroyed or directly and indirectly affected.
-crops and woods: total area of cultivated or pastoral land or woods destroyed or effects, in hectares.
-livestock: the number of 4-legged animals lost.
-government buildings: number of government and administrative buildings directly damaged or destroyed belonging to national, regional or local government.
-industrial facilities: number of manufacturing and industrial facilities directly affected.
-commercial facilities: number of individual commercial establishments (stores, warehouses, etc.) damaged or destroyed.
-transportation: the length in km of damaged/destroyed roads and railways, number of damaged/destroyed bridges, airports, marine ports.

The economic loss indicators are [5]:
-direct loss: the monetary damage of physical damage to capital and tangible wealth assets.
-indirect loss: lower output from damaged or destroyed assets and infrastructure and loss of earnings due to damage to transport infrastructure such as roads and ports, including business interruption.
-intangible costs: costs that accrue to assets without an obvious market price (difficult to depict in monetary terms).
Although there are a number of differences in terms of the organisation of the damage data within our study for the UK Environment Agency and the European &RPPLVVLRQ ¶V JXLGHOLQHV many of the data suggested for inclusion by the European Commission have been collected and reported in our study.There are some gaps, for example, there were limited data on the number of 4-legged animals that were lost.However, this may be largely due to the timing of the floods.As these were in winter, many areas that would be used for livestock grazing in summer were not populated with livestock.Table 5 sets out the results as they would be presented using the European Commission guidelines, showing that our study for the UK Environment Agency would provide all the damage data requested for indicators in [5], plus additional damages.There is slightly less granularity in the figures, for example, data were not specifically reported on the number of industrial versus commercial facilities affected.
This was a reflection of the data that were available, where information was presented on number but not type of businesses that were flooded in the numerous sources that we reviewed.All damages have been converted from Pounds Sterling (£) to Euros using a conversion rate of ZLWK GDPDJHV LQ ¼ YDOXHV. , b There was a very small number of properties that were washed away along the coast due to the coastal surge.

%HVW HVWLPDWH ¼ PLOOLRQV 5DQJH ¼ PLOOLRQV
The only indicator for which no data are available is indirectly affected people.Given that the 2013/14 floods affected large areas of England and Wales, the impacts on the road and rail systems, as well as power outages suggest that this may be a number that is much larger than the 10,465 households that were directly affected.

Comparison of the results with damages from the 2007 floods
Figure 5 presents the damages by category for the 2007 floods (left-hand figures) with the 2013/14 floods (right-hand figures).Not all categories identified in the tables above are shown in Figure 5 as the categories in 2007 were slightly different from those used in the 2013/14 study.This was to enable additional damage types that occurred in 2013/14 to be reflected and to reduce the risk of overlapping of categories and hence, the risk of double counting.
Figure 5 shows that the largest damages in 2007 were to residential properties and businesses.These were also two of the top three categories in terms of total damages from the 2013/14 floods.The overall total GDPDJHV LQ ZHUH ¼ PLOOLRQ 7KLV is 156% greater than in 2013/14.In 2007, there were 48,000 homes that were flooded during the summer months.This compares with 10,465 in 2013/14.However, the winter floods of 2013/14 saw much longer duration of flooding (especially in Somerset) and multiple sources of flooding (groundwater, pluvial, fluvial and coastal).Despite these differences, the average damages per SURSHUW\ DUH VLPLODU ¼,000 per property in 2013/14 FRPSDUHG ZLWK ¼ LQ  Damages to utilities were much greater in 2007.The approach used to value the damages from the 2013/14 floods was very similar, therefore, it can be assumed that a much larger number of people was affected in 2007.Damages to public health and welfare were also much greater in 2007.This is because the 2007 floods resulted in loss of life; this was fortunately not seen as a direct consequence of flooding in 2013/14.Total damages to transport were very similar in 2007 and 2013/ ERWK DURXQG ¼ PLOOLRQ +RZHYHU when the transport damages are considered in detail, the impacts do vary.In 2007, 84% of the transport damages were to roads.In 2013/14, this reduced to 60% with 38% of damages to rail.This included loss of the Great Western main line through Dawlish due to erosion of the railway line along the coast.This event alone resulted in 7,500 full or partial cancellations [6].