Nature-Based Solutions and Flood Risk: What the 2026 NPPF Now Requires

Nature-based solutions have moved from advocacy into planning policy. The National Planning Policy Framework published on 17 August 2026 defines them, asks development plans to make as much use of natural flood management as possible, and asks development proposals themselves to use nature-based solutions for surface water management. That came three and a half months after the Parliamentary Office of Science and Technology reported that research into the benefits of nature-based solutions outnumbers research into the disbenefits by roughly ten to one.

Both things are now true at once. English planning policy expects nature-based solutions on development sites, and the UK evidence base is not yet strong enough to tell you how much flood risk any particular feature will take away. That gap is where applications stall. It is also why a scheme leaning on performance figures borrowed from somewhere else tends to draw a request for clarification from the Lead Local Flood Authority rather than a recommendation to approve.

Research on the potential benefits of nature-based solutions outnumbers research on the potential disbenefits by roughly ten to one, according to POSTnote 768.

What follows sets out what nature-based solutions are in planning terms and where the 2026 Framework names them. It then looks at what the UK evidence will actually support, and what a drainage strategy or flood risk assessment has to do to survive consultation when nature-based features form part of the scheme.

What are nature-based solutions?

Nature-based solutions use natural processes, rather than engineered structures alone, to manage water, carbon and habitat. The 2026 NPPF glossary defines them as actions "which use natural processes to protect, sustainably manage, and restore natural and modified ecosystems that address societal challenges, benefiting both people and nature, such as the creation of woodlands or wetlands". In flood and drainage work that covers peatland restoration, woodland creation, soil and land management, leaky barriers, floodplain reconnection, field-corner bunds, wetlands, swales and rain gardens.

That wording follows the definition used internationally by the IUCN Global Standard for Nature-based Solutions, which the Environment Agency has also adopted, and it is deliberately wide. The Government's own guidance for landowners and community groups uses the same framing.

The term is broad, which is a good part of why it causes trouble. A leaky barrier in an upland headwater and a rain garden outside a block of flats are both nature-based solutions, yet they sit in different parts of the planning system, are consented in different ways, and carry entirely different weight in a flood risk assessment. One of them is a scheme somebody else built. The other is yours.

The distinction that matters is scale. Catchment-scale work is usually delivered upstream by a risk management authority, a water company or a landowner, over a period of years and with no reference to your red line. Site-scale work is yours, forms part of the drainage design, and is assessed as sustainable drainage. Both are nature-based solutions. Only one of them is a submission item.

• Catchment scale. Peatland restoration, woodland creation, floodplain reconnection, leaky barriers, field-corner bunds. Delivered upstream, usually by others, over years rather than months.
• Site scale. Swales, rain gardens, basins, ponds, wetlands, blue-green roofs and vegetated treatment trains. Delivered by the developer as part of the drainage strategy.
• Land management. Soil structure, cover crops, contour ploughing, buffer strips. Slow to take effect and dependent on continued participation by the people farming the land.
• Coastal. Saltmarsh and mudflat management, dune management and reefs, the last of which the Environment Agency added to its evidence directory only in the 2025 update.

Nature-based solutions, natural flood management and SuDS: what is the difference?

These are three overlapping terms doing different jobs. Nature-based solutions is the umbrella. Natural flood management is the subset that manages flood and coastal erosion risk at catchment scale. Sustainable drainage is the site-scale, regulated deliverable your application is actually assessed against. The 2026 Framework defines all three separately for the first time, which makes the boundaries much easier to argue in a consultation response.

Natural flood management appears in the glossary as "managing flood and coastal erosion risk by protecting, restoring and emulating the natural 'regulating' function of catchments, rivers, floodplains and coasts". Sustainable drainage appears as a system that "controls surface water run off close to where it falls, combining a mixture of built and nature-based techniques to mimic natural drainage as closely as possible, and accounting for the predicted impacts of climate change".

That second definition rewards a second reading. Sustainable drainage is now described in national policy as a mixture of built and nature-based techniques. Nature-based drainage is not an alternative to SuDS and it is not an upgrade to SuDS. It is part of what the term already means, and it has been sitting near the top of the SuDS hierarchy for years.

Upstream natural flood management never removes the need for a site-specific flood risk assessment or a compliant drainage strategy.

Keeping the three straight matters commercially, because they are routinely conflated in pre-application correspondence. An officer asking for "nature-based solutions" on a site in Flood Zone 1 is almost always asking for nature-based SuDS, and that belongs in the surface water drainage strategy for the scheme. An officer citing an upstream catchment programme as grounds to reduce attenuation on your site is asking for something no Lead Local Flood Authority is going to sign.

Where the 2026 NPPF now requires nature-based solutions

Nature-based solutions appear in four places in the August 2026 Framework, and one of them is a national decision-making policy that bites on your application. Policy N2(1)(e) asks development proposals to use green infrastructure provided as part of the scheme and nature-based solutions "to secure multiple benefits: such as for biodiversity; surface water and pollution management (including maintaining flow rates and water quality); climate change mitigation and adaptation, and recreation". That is the sentence that changes how a drainage strategy is written.

• Policy N2(1)(e), decision-making. Development proposals should use green infrastructure and nature-based solutions to secure multiple benefits, surface water management among them. This is the one that applies to a live application.
• Policy F2(1)(c), plan-making. Plans should use new development and improvements in green infrastructure to reduce the causes and impacts of flooding, "making as much use as possible of natural flood management techniques as part of an integrated approach to flood risk management".
• Policy CC1(1)(d), plan-making. Plans should identify opportunities for green infrastructure provision and nature-based solutions that safeguard carbon storage and support nature recovery and resilience.
• Policy N1(1)(b), plan-making. Plans should identify opportunities for conservation, enhancement and recovery "including through habitat restoration, the use of nature-based solutions", at catchment or landscape scale where that is what the evidence points to.

Policy F8 is where most schemes will feel the change. It asks any development proposal that could affect drainage on or around the site to build in sustainable drainage, to control flow rates and cut runoff volumes. Those systems must be designed in accordance with the National Standards for Sustainable Drainage Systems. They must have maintenance arrangements covering the anticipated lifetime of the development. On major development they must also take account of Lead Local Flood Authority advice. The Framework's national validation list asks for a statement showing how the national SuDS standards have been achieved.

Policy F8 turns the 2025 National Standards from non-statutory guidance into a matter of national planning policy.

That is a real change of status rather than a change of wording. Schedule 3 of the Flood and Water Management Act 2010 would have created a SuDS Approving Body and a mandatory approval regime. It is still not commenced in England, and the Government restated its preference in January 2026 for improving the planning-led route instead. The 2025 National Standards were published as non-statutory guidance. Policy F8 does not commence Schedule 3. It does make compliance with those standards a matter of national policy rather than local preference, which is a shorter route to the same practical outcome. The wider August 2026 rewrite is worth reading alongside it. So is the Land Use Framework published in March, which set the same direction at landscape scale. It also puts a number on the problem: about 4.6 million properties in England are at risk of surface water flooding, rising to one in four properties at risk from some source by 2050 under a high emissions scenario.

Do nature-based solutions actually reduce flood risk?

Yes, measurably, in small catchments and for frequent storms. The evidence thins quickly as catchments get larger and storms get rarer, and very little of it runs long enough to satisfy a regulator. The Environment Agency's own position, in the 2025 update to its Working with Natural Processes evidence directory, is that "it is often not possible to guarantee that NFM measures alone will provide a specified level of flood risk reduction even though they do enhance wider flood and coastal resilience".

The strongest UK dataset is Eddleston Water in the Scottish Borders, a 69 square kilometre catchment monitored for more than a decade. Field-corner bunds there attenuated discharge peaks for every storm event recorded, with reductions of up to 55%, and more than 40% of the total storage volume remained available throughout all events. At Holnicote on Exmoor, a mix of wood dams, leaky weirs, upland attenuation ponds, woodland creation and floodplain reconnection may have reduced flood peaks by up to 38% in major events. On the River Otter, two beaver families built 28 dams within four years, slowing and backing up nearly two kilometres of river. In Norfolk, the Wendling Beck partnership has restored around 800 hectares of arable land to mixed habitat. Those sit alongside the wider set of delivered UK natural flood management schemes at Calderdale, Pickering and elsewhere.

Eddleston Water is one of very few long-term, monitored, real-world UK datasets on natural flood management, and it covers a single catchment of 69 square kilometres.

The qualifiers around those numbers do a great deal of work. The Environment Agency's 2025 evidence update reviewed roughly 800 studies across 17 natural flood management measures and publishes no general "natural flood management reduces peak flows by X%" figure, because the effect is site-specific and confidence varies by measure. Where it does offer a threshold, it is a modest one: leaky barriers can generally show flood risk reduction for small floods in small catchments, below about 10 square kilometres.

None of that makes them ineffective. It makes them hard to quantify at the resolution a planning submission needs, which is a different problem, and one to handle in the design rather than argue away in a covering letter.

Why does the evidence lag the policy?

Because monitoring costs money that nobody is committing for long enough. POSTnote 768, published on 1 May 2026 by Jonathan Wentworth and Ben Cutting, describes a loop. Proving catchment-scale nature-based solutions work needs long-term, large-catchment trials. Funding long-term, large-catchment trials needs funders willing to commit before that evidence exists. People who gave evidence to the briefing called the current funding too siloed between sectors, too short-term, too little, and too narrow to see a whole catchment.

The numbers bear that out. Defra's dedicated natural flood management programme is worth £25 million and runs to 31 March 2027, supporting 35 active projects. The wider flood and coastal settlement announced to Parliament in March 2026 is £1.4 billion for 2026/27 alone, £4.2 billion over three years and £7.9 billion of capital across ten. The dedicated evidence programme is a rounding error against the delivery budget it is meant to inform.

There is a distributional problem underneath it as well. POSTnote 768 notes that market-based instruments for funding nature-based flood risk work "can disproportionately benefit affluent areas", because landowners with better access to finance are likelier to host schemes, while more deprived catchments further downstream carry more of the exposure. So the monitoring data builds up where the schemes are, not where the risk is, and the literature slowly comes to describe a set of catchments that are easier to work in and cheaper to buy into rather than the ones carrying the exposure, which is a quiet form of selection bias that nobody set out to create.

Drought is worse served again. The briefing records that far less research has been done on drought than on flooding, because floods are more visible and arrive faster. Given that the same catchment can now be too dry in one season and overwhelmed in the next, and that hard-baked ground sheds summer rain straight into the drainage network, that is a gap worth watching. The 2025 National Framework for Water Resources asks the regional planning groups to coordinate local uptake of nature-based solutions where they deliver value, which is a reasonable ambition resting on a thin base. The February 2026 revision of the Treasury's Green Book pushes the same way, requiring appraisals to reflect the environmental principles policy statement and the biodiversity duty from the longlist stage onwards.

What does this mean for a live planning application?

Nature-based features belong in your drainage strategy as designed, sized, maintainable elements, with engineered fallbacks carrying the residual risk. Performance figures from other catchments carry no weight without site-specific justification. Climate change allowances still apply in full. An upstream catchment programme does not reduce what your site has to attenuate.

Three things go wrong most often, and all three are avoidable at concept stage rather than at appeal.

The first is borrowed numbers. A 38% peak reduction at Holnicote is a real result for that scheme on those storms. It is not a factor that transfers to a different catchment, a different soil, a different rainfall profile or a different feature. A drainage strategy leaning on published performance data without site-specific hydraulic modelling will draw a holding objection, and rightly so.

The second is climate change allowances. The Environment Agency's allowances are derived from UKCP18 projections of peak rainfall intensity and peak river flow. They are not calibrated to the assumed performance of a vegetated feature with a short monitoring record. Where nature-based features form part of the mitigation, the residual risk assessment has to assume they underperform, particularly given how much rainfall intensity has already shifted.

The third is framing. POSTnote 768 records the view that nature-based solutions should be presented "not as a cheaper alternative but as part of a catchment-wide approach integrated with grey infrastructure". The Environment Agency's position statement puts it plainly, and the same logic holds at site level. A scheme offering a wetland instead of attenuation is tested far harder than one offering a wetland as part of the attenuation. The word that matters is and, not or.

What a Lead Local Flood Authority will expect to see

The Lead Local Flood Authority is the statutory consultee on surface water for major development, and Policy F8 now says so directly. Nature-based features do not change what it asks for, and they do not exempt a scheme from the design conventions in the CIRIA SuDS Manual (C753). They change how much of it you have to evidence.

1. Discharge rates and volumes calculated conventionally. Greenfield runoff rates, long-term storage and volume control derived the normal way. A nature-based feature is credited for the storage and interception it can be shown to provide, not for the concept.
2. Storage demonstrated, not assumed. Every attenuating feature sized in the model, with its contribution set out in the calculations. Vegetated features are often over-credited on paper and under-perform in silt-loaded reality.
3. Infiltration evidenced where it is claimed. BRE 365 testing at the right depth and season, with the groundwater level established. A rain garden on clay is landscaping, not drainage, and calling it drainage in a submission is the fastest way to lose a fortnight.
4. Exceedance routed above ground. Exceedance flow routing matters more, not less, where the primary system depends on vegetation and permeability that can degrade.
5. A maintenance regime with a named owner. Cutting, silt removal, replanting, inlet and outlet clearance, with the funding route stated. Policy F8 asks for arrangements covering the anticipated lifetime of the development.

The commonest reason a nature-based drainage proposal is queried is not the principle. It is that the storage is asserted rather than calculated.

Where a proposal is properly evidenced, the nature-based element usually helps. It answers water quality, amenity and biodiversity at the same time as quantity. Where it is not, it produces an LLFA objection that costs more time than the design ever saved, and a flood risk assessment for planning that has to be reworked. That is an expensive way to be green.

When are nature-based solutions the wrong answer?

When the ground, the space or the storm will not support them. Nature-based drainage depends on infiltration, land take and vegetation health, and there are common site conditions where at least one of those is missing. Saying so at concept stage is far cheaper than discovering it at condition-discharge stage.

• Low-permeability ground. Clay, or made ground over clay, will not sustain infiltration. A vegetated feature can still attenuate and treat, but it cannot dispose.
• High groundwater. Where the water table sits close to the surface for part of the year the unsaturated depth beneath a feature disappears, and so does the storage you sized on it.
• Contaminated land. Infiltration through a contaminated profile risks mobilising the contamination into groundwater, and the regulator will say so.
• Tight urban sites. Nature-based features need surface area. On a constrained brownfield plot the land take is often the deciding constraint, whatever the policy preference.
• The largest events. Effect falls away as storms get rarer, which is precisely the return period a flood risk assessment is designed around.

Two of these run against the common assumptions and deserve more than a line each.

Peatland restoration is frequently described as a flood measure. POSTnote 768 is careful about it: undisturbed peatlands are typically saturated or close to saturated throughout the year, so they do not attenuate floods by storing extra rainfall in the soil. Their value is carbon, water quality and habitat — only 22% of UK peatlands are near-natural and storing rather than emitting carbon, and under 13% in England — with flow benefits that are real but secondary and hard to attribute.

Trees are the other. Rainfall interception falls away as a storm grows larger and more intense, and forest soils are often already saturated during heavy winter rainfall. Woodland does least in exactly the events an assessment is sized for. That is an awkward fact and it is not going away. There is a genuine conflict between the two halves of the resilience argument as well: POSTnote 768 records that nature-based solutions increasing evapotranspiration up-catchment may reduce water availability down-catchment in drought conditions. The measure that helps in February can hurt in August.

Who pays for it, and who adopts it?

On a development site, you pay for it, and the adoption route decides whether the feature survives its first decade. Nature-based drainage carries a higher maintenance burden than a buried tank, and far less tolerance for neglect. That is why Policy F8's requirement for maintenance arrangements covering the anticipated lifetime of the development is not a formality.

Vegetated features need active management: silt removal, cutting regimes, replanting, inlet and outlet clearance. Unlike a pipe, they show it when that stops. A blocked outlet is visible from the pavement. Adoption is the recurring difficulty. A water company may decline a feature it does not consider adoptable. A local authority may decline one it cannot resource. A management company arrangement has to be funded for the life of the scheme rather than the life of the sales programme, and that commitment has to be visible at the point the drainage condition is discharged.

A nature-based feature nobody has agreed to maintain is a planning condition waiting to be breached.

None of this argues against nature-based solutions. It argues for treating them as engineering rather than as gesture. The 2026 Framework has asked for them, and the evidence base will catch up unevenly over the next decade. In the meantime the defensible position is the one POST and the Environment Agency both land on. Design nature-based features as part of an integrated system, and size the engineered element for what happens when the vegetation is three years old, silted and halfway through a wet January. If you want that written properly into a drainage strategy or a flood risk assessment, our consultants do it every week.

Frequently asked questions