Understanding Flood Risk
In England there are around 5.2 million properties at risk of flooding. If that’s not alarming enough, around two-thirds (67%) do not understand their flood risk. With many properties at risk from multiple sources of flooding and the average UK household more likely to get flooded than burgled, we need to make the complex topic of flood risk easy to understand.
This blog focuses on river flooding events; also known as ‘fluvial’ events. They have the largest impacts on communities, gain the greatest media attention and often use return periods to refer to floods as a “1 in ‘X amount’-year” event. But what does this actually mean and why is the use of return periods not the best means of communicating flood risk?
Also known as recurrence intervals, return periods are the product of frequency analyses. When calculating return periods for river flooding, data from river level monitoring stations is used to plot the volumetric flow rates within a river channel; with the rate of flow usually measured in cubic meters per second (m3/s). Statistical analysis is used to estimate the frequency with which certain river levels and flow rates are likely to re-occur.
Hydrologists use computer modelling to simulate different rainfall intensities and durations over a catchment in order to see how the flow rates within a river are affected. Lots of variables factor into this process, including the permeability of different land surfaces and run-off rates, flow paths through the catchment as well as the characteristics of the channel itself and climate change projections.
The important thing to know is that as the sampled data pool grows with each new flood event that occurs, return periods for a particular magnitude of event can change. This is why you may hear a recent flood event referred to as a ‘1 in 100-year’ event, and later re-categorised as a ‘1 in 80-year’ event. Return periods are not an exact science and a ‘1 in 100-year’ flood event can occur once, twice or more, or not at all, within a 100 year timeframe.
When return periods are quoted on television reports and in newspaper articles, they give the impression that these events are cyclical, and that therefore we can predict them. The public is presented with the ‘1 in 100-year’ sound bite and forgiven for thinking that now the flood event has occurred, the next one is unlikely to occur for another 100 years. In fact, there is a 63.4% chance of one or more ‘1 in 100-year’ flood events occurring in any 100 year period!
A better way to express risk
Rather than state return periods, a better way to communicate risk is to convert the return period to show the probability of a flood event of a certain magnitude or greater, occurring in each and any one year and to express this figure as a percentage. This method is useful because it helps people to visualise risk more clearly in their minds by expressing it in a more relatable and quantifiable way. This is known as the ‘Annual Exceedance Probability’ (AEP); the inverse of the return period, and is expressed as ‘%AEP’. As can be seen from the table below, a flood event with a return period of 100 years (a ‘1 in 100-year’ event) would have a 1% chance of occurring each year. And a flood event with a return period of 2 years (a ‘1 in 2-year’ event) would have a 50% chance of occurring each year.
To help express risk in more simplified terms, The Environment Agency has assigned risk bands to the spread of %AEP. They also provide online flood risk maps which allow users to locate their property within the floodplain of a river. As can be seen from the table below, %AEP is spread across four ‘risk’ categories; Very Low, Low, Medium and High.
These categories are illustrated on the flood risk map by different blue coloured bands which are overlaid across the floodplain; the darker the shade of blue, the higher the risk. You can check your flood risk online by entering your property number and postcode here: https://flood-warning-information.service.gov.uk/long-term-flood-risk/map
What’s important to remember is that while annual exceedance probability is used to communicate risk in a different way, %AEP calculations are still derived from return periods!
Grouping the spread of %AEP to within the four categories above, effectively seeks to express flood risk using four broad brushstrokes; which works well because it allows the user to easily know their level of risk. But this method of communicating risk also brings with it the following limitation… While the user ‘knows’ what their flood risk is, they are unable to put that risk into any kind of context to help them fully understand it; and therefore cannot determine whether that particular level of risk is acceptable to them, or not.
Putting risk into context
If you want to understand your flood risk in more detail, you would be advised to obtain a site-specific flood risk assessment for your property, produced in consultation with a professional. The flood risk specialist can utilise detailed data sets and maps to assess risk from different sources and determine expected flood depths across a range of return periods. They can also physically survey the site to assess property specific flood risk attributes such as elevation. In some cases, residents have successfully disputed their level of flood risk by having a flood risk assessment carried out for their property, and demonstrated to insurers that their risk is actually lower than stated. The four risk bands give you a broad picture, but they are not designed to present you with detail.
There is another thing to consider! With ‘annual’ exceedance probability, each year is effectively a new role of a dice. For the ‘1 in 100-year’ event, a 100 sided die (known as a Polyhedral die) would be used and a flood would be indicated by say, rolling a number 5. With each roll, the chance of rolling a 5 (a flood) remains the same; a 1 in 100 chance, or 1%. The probability of a flood remains the same for each consecutive year the die is rolled. However, the more times you roll the die, the chances of hitting a number 5 are inevitably increased!
Time and the cumulative factor
The vast majority of people will not live at a property for just one year; especially if they are buying it as opposed to renting. In determining whether a property’s level of flood risk is acceptable or not, it is useful to consider the duration of time that you expect to live there. While ‘annual’ exceedance probability looks at the risk of flooding annually, it may also be worth knowing your chances of experiencing a flood across a longer time frame, such as the 10, 20 or even 30 year period of time that you may reside at the property. The results are surprising!
As can be seen from the table below, if you’re buying a property with a 0.1%AEP (the bottom end of the ‘Low’ risk band) and expect to live there for only 1 year, the chance of a flood is 0.1%. If you expect to live there for 20 years, the probability of you experiencing a flood within that time period increases to just short of 2%; still relatively low indeed. In contrast however, if your property falls at the top end of the ‘Low’ risk band at 0.99%AEP; your chances of experiencing a flood over the same 20-year period would be 18.04%! While a 0.99% chance of a flood occurring each year sounds low, if you knew that there was nearly a 1-in-5 chance you’d experience a flood at some point within the next 20 years, would you still buy that property?
For some property owners, a 2% chance of flooding may be an acceptable level of risk to take over a 20 year period, whereas an 18% chance, close to 1 in 5, would not. Two very different flood risk outcomes in terms of probability and both falling within the same ‘Low’ risk category.
The understanding of flood risk could prove to be the greatest driver and motivator in encouraging people to become more flood resilient. If a property’s level of flood risk is unacceptable, the owner can make the most informed decision on whether to spend money protecting their home, or move house to a different location.
The online flood risk maps currently available to the public are a great resource which should be used by everyone at risk of flooding. They act as a good ‘indicator’ to determine whether a more detailed view of their flood risk is required; and if so, a flood risk assessment can be undertaken. At some point in the future, online flood risk maps will likely make use of more granular data and updates will bring higher resolution maps with more detail; allowing the user to select their property on the map and obtain a more defined %AEP of flood risk from rivers and the sea. The provision of a simple box whereby the user can enter the number of years they are likely to reside at the property could provide the calculations required to help put their flood risk into context over the longer-term.
Although we can use tools such as these to help to express flood risk in a more relatable way, the uncertainty of return period calculations cannot be escaped. Hydrology is not an exact science and it’s important to remember that flood events of any magnitude can occur at any time. Because of this, when it comes to communicating flood risk to aid understanding, there remains as of yet, no better means than conversation and explanation.