What is striking in these responses is the extent to which respondents drew on their own experience and judgement or on other informal sources such as neighbours rather than on formal sources such as the Environment Agency’s letters, website or flood warnings, the local council, water companies and solicitors’ searches. When properties are bought and sold in England and Wales, part of the normal legal process is a solicitors’ search on behalf of a potential buyer of legal documents relating to the property and this should include enquiries to check with the Environment Agency and local authority as to whether there is any history of flooding or evidence of flood risk at the property. In the Lower Thames Survey this legal mechanism did not appear to be an important source of flood risk information. Currently, a seller is under no obligation to reveal a known flood risk although under proposed new legislation (to require sellers of property to provide a Home Information Pack - a set of standard information about the property) this might change.
In the Lower Thames, those who believed their home to be at some risk of flooding either in the five year period or in the 50 year period, were asked ‘Do you think you would have moved to the area if you had known about the possibility of flooding?’ All those with awareness of the flood risk prior to moving, responded positively. Of the small group (48) who were not aware when they moved to their home, the majority 58% said they would have moved had they known of the risk, 21% would not have moved and the same proportion did not know. Overall, 81% of those with some current awareness of the risk stated that they would have moved even if they had known of the risk and only 10% would not have moved. For respondents having at some time had some experience of flooding (not necessarily inside the home), living on the river bank and social class did not have any influence on the decision to move despite the flood risk. These findings can be taken as illustrating the strong pull of the amenity of living in the Lower Thames area balanced against their limited knowledge of the risk of flooding and also their limited experience and thence, possibly, their poor understanding of the impacts that flooding would have on their lives and property (McCarthy et al 2006).
Another factor which may affect the level of flood impacts and hence respondents’ coping capacity are flood warnings.
Flood warnings authorities
In England and Wales the Environment Agency has (since 1996) had responsibility for issuing flood warnings, a responsibility previously exercised mainly by the police. The Agency is also (along with the emergency services and local authorities) a Category 1 responder under the Civil Contingencies Act (2004) providing civil protection during an emergency situation, such as a flood event. The flood warnings issued by the EA only apply to river, tidal, and coastal flooding, and not to flooding from other causes such as surface water, sewers and drains (although this is currently being considered). The main warning dissemination methods currently in use in England and Wales are: automatic voice messaging systems (AVM, now Floodline Warnings Direct or FWD) using land-line and mobile telephones, faxes, sirens, loudspeakers, face-to-face door knocking, written communication, flood wardens, TV weather reports, teletex and radio. Other methods soon to be available include SMS texts to mobile telephones and digital broadcasting.
As part of their role as Category 1 Responders under the Civil Contingencies Act, Local Authorities also have responsibility for providing an immediate response to care for flood-affected populations, including the provision of emergency care, feeding, accommodation and welfare of evacuees. They also have the most responsibility following flooding in assisting local communities during the recovery phase.
Receipt of flood warnings
In September 1996, prior to all the surveys and flood events analysed for this report, the EA took over lead responsibility for disseminating flood warnings to the public. When the flooding occurred in Easter 1998, the Agency had only just begun to make improvements to the warning system and its failure to provide warnings to many of those affected by that event was severely criticised (Bye and Horner, 1998). Since that time, the EA has focused attention on enhancing its flood forecasting, and especially its warning systems. It has expanded its ability to deliver warning messages directly to the public: first through its AVM system, and since early 2006 through Floodline Warnings Direct, a national automatic telephone warning system with greatly increased capacity to deliver warning messages (Andryszewski et al., 2005). The Agency offers different levels of service to different areas according to the level of risk, from a maximum of direct warning to properties, to a minimum in which only the media and professional partners such as the local authority and emergency services are used to broadcast warnings, with an intermediate category in which sirens or loudhailers are used in the community. Flood warnings are provided by the EA to known areas of flood risk but flooding can occur in areas where no flood warning dissemination service is provided. For some events such as those affecting very small watercourses and very extreme events, no warning service is provided.
The surveys included in this report cover flood events occurring over the period in which the Agency has been active in improving its forecasting and warnings service. All the surveys provide some evidence on the receipt of warnings. In the surveys, respondents were left to define what constituted a warning and their responses cover warnings deriving directly from official sources, informal warnings and indeed their own experience and judgement. The data show that informal sources of warning such as family, friends and neighbours were an important source of warning. These warnings may serve to amplify, supplement or indeed compete with formal systems (Parker and Handmer, 1998) (Table 4.7).
Table 4.7: Flood warnings received
Intangibles: flooded (a)
Lower Thames (c)
No warning received/received too late
Don’t know/no answer
Warning lead time: those receiving a warning and aware of lead time
EA automatic voice message (AVM)/ recorded telephone message from EA
Environment Agency Floodline
Personnel/telephone call from/to EA
Emergency services (Fire/Police/Ambulance)
) 24 (35)
(a) Respondents were asked ‘Did you receive a warning from any source before the flood (most recent/worst flood)?
(b) Respondents were asked about a warning received in the most recent/worst flood.
(c) Respondents were asked: ‘Have you ever received a warning from any source, even if you haven’t been flooded?’
(d) Respondents were able to name more than one source of warning received in the worst/recent flood.
The data suggest that there may have been some improvement in flood warning dissemination over time. In the Intangibles Survey, only 10% of residents flooded in 1998 reported receiving a warning; for 1999 and 2000, the proportions were 29% and 28% respectively. The Warnings Survey, however, does not provide evidence to confirm an improving trend since those interviewed in the first Phase of this survey who had been affected by flooding in 2003/4 were no more likely to have received a warning than those affected by earlier events covered in Phase 2 (32% warned in Phase 1 compared with 41% in Phase 2). A key finding from the surveys analysed here and from the research undertaken for the Environment Agency (Tunstall et al., 2005) is that despite investment in improvements to warning systems, it is still only a minority of residents (rarely more than 40%) who receive a flood warning of any kind.
The data from both the Intangibles and the Warnings Surveys also indicate that the receipt of a warning is very area and event specific. For example, in the Intangibles Survey there were ten locations in which no one reported receiving a warning. These included Banbury, flooded in the Easter 1998 event, and Waltham Abbey, flooded in autumn 2000. In five locations, a majority of residents were warned (RPA/FHRC, 2004). A similar pattern of wide variations in the receipt of warnings according to location was observed in the Warnings Survey (Tunstall et al., 2005).
4.4.5 Who receives a warning?
The Environment Agency’s AVM system (now Floodline Warnings Direct) has become the main method by which residents in England and Wales are warned. The growth in importance of the AVM (now FWD) system may be reflected in the sources of warning found in the Intangibles and the Warnings data sets, since the Warnings Survey was focused on more recent flood events than the Intangibles study (Table 4.6). In the current ‘opt-in’ system residents have to take action to register their household on the system in order to receive warnings in this way, although an ‘opt-out’ system is currently being considered whereby residents will automatically be registered on the system unless they request otherwise.
In the Warnings Survey, being registered on the AVM system was found to be the main factor associated with receiving a warning. Overall, 51% of respondents who answered the question in the survey were registered on the system at the time of the survey. This proportion includes some who signed on to the system after the most recent flood. All registered residents were twice as likely to have received a warning as those not registered (48% compared with 24%). In Phase 2 of the Warnings Survey, a distinction was made between those registered at the time of the flood event (only 28% of Phase 2 respondents) and those who registered afterwards (24%). In this Phase, three-quarters (76%) of those registered, compared with only one-third (27%) of those not on the system at the time of the flood, received a warning. The interesting question then is who registers to receive the AVM/FWD and why do they do so? These questions are examined in Section 5.2.1 on preparedness.
Other factors were found to be associated with receiving a warning in the analysis of the Warning Survey data (Tunstall et al., 2005). Although the association between length of residence and the receipt of a warning was not strong, more long term (those resident 20 years or more) than recent residents had received some kind of warning. This did not appear to be due to the long term residents being more likely to be signed up to the AVM system. It may be because long term residents have greater experience and awareness of the flood risk. Certainly, those residents who reported that they were aware of the flood risk prior to the recent flooding were significantly more likely to have received a warning than those unaware (47% compared with 23%). Again this did not appear to be simply due to a higher proportion of those with prior awareness being registered on the AVM. Not surprisingly, those with prior experience of flooding inside their homes above floor level were not only more likely to have been registered on the AVM at the time of the recent flood, but also to have received a warning. The proportion receiving a warning was 54% for those with some past experience of in-house flooding, compared with 29% for those flooded in their house for the first time.
In the Warnings data set, there was no evidence of warnings being targeted at those most at risk and thence most affected. Those who had flood waters inside their homes were no more likely to have received a warning than those less seriously affected.
New analyses of the Intangibles data set have been undertaken to see whether this data set, which contains a larger number (albeit a smaller proportion) of respondents who actually received a flood warning, can throw further light on the factors affecting receipt of a warning. The nature of the flood event can make a difference to the possibility of forecasting and issuing timely warnings. Indeed, some events such as the flooding in Boscastle in 2004 (Environment Agency, 2005), are regarded as extremely difficult to predict with current science and technology. In the Intangibles Survey, respondents were asked a question that was not included in the Warnings Survey: ‘How quickly did the floodwaters rise’? Their perceptions of the speed of onset of the flooding were associated with the receipt of a flood warning and with the length of warning lead time. Those who reported the waters rising quickly were less likely to have received a warning and have a shorter warning time than those who judged their flood to be slower in its onset (Table 4.8).
Table 4.8: Receipt of a warning and warning lead time according to the perceived speed of onset of flooding. Intangibles Survey (RPA et al., 2004)
Warning lead time: those in receipt of a warning and aware of the warning lead time %
Under 2 hours
2 hours< 4 hours
4 hours < 8 hours
8 hours or more
Number of cases
* Chi square; p< 0.001
**Chi square : p< 0.001
All the respondents in the Intangibles Survey had flood waters inside their homes. However, there was some variation in the extent to which their properties were affected. There were four types of main parts or rooms affected by flood waters: living rooms, bedrooms, kitchens and bathrooms - all taken as a measure of the extent of flooding. A majority in this survey had had two parts affected, usually the kitchen and the living room, since most of the properties were houses on two floors. However, there were some respondents who did not have any main rooms affected because the flood waters only entered other parts such as hall ways, cellars or basements. It was found that those more seriously affected were more likely to have received a flood warning than those with fewer main parts of their home affected. In addition, there was an association between depth of flooding and receipt of a warning (Figures 4.1 and 4.2).
This is surprising since the Environment Agency is not able in most areas to target its warnings specifically at the properties most at risk. It may be that:
those properties that have flooded before, and are most at risk, are known to the warning authorities; or
people resident in such properties have taken steps to ensure that they are registered to receive warnings; or
people resident in such properties have taken steps to contact the warning authorities themselves as flooding threatened.
Figure 4.1: Receipt of a flood warning by number of main rooms flooded: Intangibles Survey
Chi square; p< 0.05
Figure 4.2: Receipt of a flood warning by depth of main room flooding: Intangibles Survey
Chi square; p<0.01
Certainly, in the Intangibles Survey (as in the Warnings Survey), those who were aware i.e. who reported that they were aware of the flood risk in the area before they were first flooded, were more likely to have received a flood warning than those unaware (39% compared with 20%). Most of those interviewed in the Intangibles Survey (80%) had never before been flooded inside their home at their current address. Those that had previous experience of such flooding (mainly of one earlier flood inside their home) were significantly more likely to have received a flood warning than those experiencing flooding for the first time in the recent or last event (31% compared with 23%).
The social characteristics of those receiving a flood warning in the two surveys were examined to see whether any common factors could be identified. In both surveys some differences were found according to social grade. These differences depended in part on the social grade groupings used. However, significant differences were noted, in both data sets, with the categorisation used in Figure 4.3. In the Warnings Survey both the skilled, semi-skilled and unskilled worker categories were less likely to have received a warning. In the Intangibles Survey, the lowest social grade was the only one to be markedly different.
There was some evidence of differences in the receipt of a warning according to tenure. In the Warnings Survey, more owner-occupiers than those renting property were warned (39% compared with 23%). However, differences were not statistically significant and the number of respondents renting property (39) was very small. In the Intangibles Survey, both those owning and buying their properties were more likely to have received a warning than those in other tenure groups, a substantial number (115) in this survey. The proportions were 23% and 26% respectively for those owning and buying and 13% for those in the other tenure groups. It is not possible to establish whether these differences were due to variations in service provision to different neighbourhoods and groups or to differences in the extent to which residents are proactive in accessing services.
No other differences were found in the two data sets in the receipt of warnings according to other social characteristics. Households that are vulnerable in terms of age, disability or ill health of their members, and those living alone and households with children under the age of ten, were no more likely to have received a warning than less vulnerable households. As yet, the Environment Agency’s formal warning systems generally are not able to differentiate and prioritise particularly vulnerable households, although this may be done in certain locations and circumstances, for example, where flood wardens or informal warnings systems and active neighbourhood networks operate.
Figure 4.3: Receipt of a flood warning by social grade