Modelling a small watercourse with no published model

On a site beside a small stream, the flood map shows a Flood Zone 2 and 3 outline generated by a national model and nothing else. There is no Environment Agency model, no modelled levels, no node data. The planning officer wants a finished floor level and the applicant wants an answer this month.

The way through is to establish flows for the catchment, check what the channel can carry, and model only as much as the decision requires. That is a defensible approach where no Environment Agency model exists, provided you say what you did and why.

Establish what data actually exists

Before assuming there is nothing, ask. The Environment Agency supplies flood risk data products on request, including detailed modelled levels and flows where a study covers the reach, historic flood records, and the underlying model data where one exists.

Request the full set for the reach and be specific about the location, because coverage can begin a few hundred metres downstream of a site. Historic flood outlines and records of past events are often available even where no model is, and on a small watercourse anecdotal flooding history held by the lead local flood authority or the parish council can be more useful than any of it.

Allow time. Data requests take weeks rather than days, and the answer may be that nothing exists, which is itself a finding to report.

Main river or ordinary watercourse

This determines who you are dealing with and which consent regime applies to any works.

Main rivers are designated and shown on the Environment Agency's main river map. Everything else is an ordinary watercourse, where the lead local flood authority or an internal drainage board is the risk management authority, and works affecting flow need land drainage consent under the Land Drainage Act 1991 rather than an environmental permit.

Most unmodelled small watercourses are ordinary watercourses. That usually means a more pragmatic consultee, a less prescriptive expectation about modelling, and a real interest in whether the scheme affects conveyance. It also means the Environment Agency may not comment at all, so do not wait for a response that is not coming.

Estimating flows for a small catchment

Delineate the catchment from the terrain data, then derive the design flows.

For catchments within the applicable range, the statistical method and the revitalised rainfall runoff method from the Flood Estimation Handbook are the standard tools, using catchment descriptors obtained from the published service. Urban extent, soil type and the presence of any attenuation upstream all change the answer, and on a small catchment they change it a lot.

Very small catchments fall outside the range the main methods were derived for, and the established small catchment guidance is the right reference there. Whatever method you use, name it, state the version, and record the catchment descriptors and any adjustments applied.

Two checks worth running. Compare the estimated flow against any gauged data on a hydrologically similar neighbouring catchment, scaled by area. And sense-check the resulting depth against what the channel visibly carries, because an estimate that puts a metre of water over a bank that shows no sign of ever overtopping is probably wrong.

A channel capacity check and when it is enough

On many small sites this is all that is needed.

Survey a representative cross-section, or better several, calculate the normal depth capacity using an appropriate roughness coefficient, and compare it with the estimated design flows. The result tells you the return period at which the channel begins to spill, and where the water goes when it does.

Structures are what actually control the answer. A culvert under a farm track, a pipe under a road, a footbridge with a low soffit: each creates a backwater and each is a blockage risk. Survey the inverts and soffits, and test the capacity with the structure clear and with it partly blocked.

A capacity check is enough where the site is not in the floodplain, the flow path is clear, and the conclusion is not marginal. It is not enough where the answer determines a finished floor level within a few hundred millimetres, or where the scheme changes the conveyance. The reasoning belongs in the Flood Risk Assessment in plain terms, so a reviewer can see where the judgement was made.

Survey data a reach model needs

If a model is required, the survey drives the cost and the quality.

• Cross-sections at intervals appropriate to the channel, closer together where the geometry changes, extending up the banks and across the floodplain to a level above the expected flood level.
• Every structure surveyed in detail: invert, soffit, span, skew, screen arrangement and any control.
• Bank top and floodplain levels, tied to Ordnance Datum, with a terrain model covering the extent of any out-of-bank flow.
• Photographs of the channel and each structure, taken at the time of survey, for roughness estimation and for the record.

A 1D model with mapped floodplain extents is usually proportionate on a small watercourse. A 1D-2D linked model is warranted where out-of-bank flow spreads across a wide floodplain or takes a route away from the channel, which on some small streams it does.

Proportionate modelling on a small site

The level of assessment should match what the decision needs. National guidance on flood risk assessment says as much, and consultees generally accept it on ordinary watercourses.

For a single dwelling on the edge of a mapped outline, a capacity check plus a level survey plus historic evidence will usually do. For a scheme of a dozen houses within the outline where levels are marginal, a reach model is proportionate. For anything that alters the channel, culverts a length of it, or places built form in the flow path, expect to model, and expect the presumption against burying a stream in a culvert to be raised.

Where the scheme is in Flood Zone 3 and vulnerable, the modelling has to be good enough to support the exception test evidence, because a marginal conclusion from a coarse model will not carry the second limb.

Agreeing the approach before you model

Write a short scoping note, one or two pages, setting out the catchment, the data available, the method proposed, the events to be assessed, the climate change allowances and the blockage scenarios. Send it to the lead local flood authority, and to the Environment Agency where the watercourse is main river.

Most authorities will respond, and a response saying the approach is acceptable is worth more than any amount of modelling done on assumptions. Where no response comes, record that the note was sent and proceed. An unanswered scoping request documented in the assessment is a reasonable position.

Presenting uncertainty honestly

An estimate on an ungauged small catchment carries real uncertainty, and pretending otherwise damages the assessment.

Report the flow estimates with the method and the confidence the method carries. Run a sensitivity test on roughness and on the blockage assumption, and show what happens to the level in each case. Where the difference between scenarios is material to the floor level, adopt the conservative one and say why.

Then state the residual risk plainly: what happens in an event larger than the design case, where the water goes, whether access and egress remain available, and what the occupants would do. That section is short and it is the part a reviewer reads first.

Coordinate the whole thing with the surface water design, because a scheme discharging to the same watercourse has to demonstrate that it is not adding to the peak at the wrong moment, which is where the Drainage Strategy and the flood assessment meet.

Enter your postcode on PlanScope to see whether a flood risk assessment is likely to be triggered for your site. This is general guidance, not legal advice; check the position with your local planning authority or a qualified professional before relying on it.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.