Not across the town, but very much at certain addresses. High ground on the Mendip edge stays dry. The flat land below was moor before anyone drained it, and drained land remembers what it used to be. A small number of properties carry real risk while the great majority carry none. Flood risk here is an address question, not a postcode question, and the difference matters more than the headline.
In the Environment Agency's river-and-sea flood zones, about 6% of the area's postcodes fall inside zone 3, the highest-probability band, and around 3% sit in zone 2, which is the additional medium-probability fringe lying beyond zone 3 rather than a larger area containing it. Around 36% of postcodes sit within the Axe Brue Internal Drainage Board area, which is the administrative acknowledgement that this landscape is kept dry by pumps and channels rather than by geography alone.
Surface water is not published here as a separate share. It is folded into the combined postcode indicators further down, alongside rivers and sea, so the figures in that measure already carry it. Anyone treating the river-and-sea zones as the whole answer on a moor-edge site is reading half the problem.
The forward look is where the numbers move. Once the Environment Agency's climate allowance is applied, the modelled extent rises to around 35% of postcodes. That is not a prediction about any individual house, and it is not a statement about next winter. It is the mapped area under a wetter set of assumptions, and it is the version a mortgage lender in twenty years will be looking at rather than the current one.
For insurance, the practical consequence is that quotes will diverge sharply between neighbouring addresses. A property inside a modelled extent can attract higher premiums, higher excesses on escape-of-water and flood claims, or conditions attached to cover. A property outside one is usually unremarkable. Get an indicative quote on the specific address before instructing a solicitor, because the postcode average tells you nothing useful about the house you are actually buying.
On the combined postcode indicators, which rate each postcode on at least one flood measure including surface water, about 8% come out high, around 9% medium and about 30% low. These bands overlap with the river-and-sea zones rather than sitting alongside them, so they are a second view of the same landscape, not an additional set of at-risk places to be added on.
Counted as homes rather than postcodes, the Environment Agency's rivers-and-sea mapping lists 992 properties across the 44 postcodes that carry any risk at all. Of those, 37 are at high risk, 26 at medium, 61 at low and 249 at very low. Only one of the high-risk properties is residential. The rest of that high band is agricultural and commercial, which is what happens when the risk sits on working moorland rather than on the streets people live on.
Whether anything stands between the water and the building is a separate question. About 6% of postcodes sit in a high-probability modelled extent with no defence present in the model, against roughly 1% where a defence is present. A defence appearing in the model is not a promise of protection: it means a structure has been accounted for, not that it will hold, not that it is maintained to a standard, and not that the property behind it is safe.
Further out, the shoreline management policy for this stretch is to hold the line over the next twenty years, move to managed realignment over fifty, and hold the line again over a hundred. Those are stated intentions about maintaining or repositioning defences, not commitments, and not evidence that a defence already exists where you are standing. Policies are reviewed, and managed realignment at the fifty-year horizon means somewhere, at some point, the line is expected to move inland.
At the nearest tide gauge, Devonport, the recorded rate of sea-level change across the 1962 to 2015 record is about 2.35 millimetres a year. That is measured history rather than forecast, and it is the background against which every defence decision above gets made.
There is no recorded historic flooding in the area on the records held here. That is genuinely reassuring, and it is not proof that water has never come up. Absence of a record means nothing was logged, not that nothing happened.
Flooding here sits in the middle: a real and mapped problem for a small number of addresses, an irrelevance for most of the rest, and a question that only gets a proper answer when you put the exact address into the Environment Agency's long-term flood-risk tool rather than the postcode.