Nature may hold the solution to future flooding

In ‘View on the Stour near Dedham’, John Constable painted a tranquil river Stour gently meandering through the Suffolk countryside towards Dedham village. On the canvas Constable illustrated idyllic rural life by the water: in the foreground a rake lies abandoned on the riverside, in the near-distance men labour to separate two barges, whilst a woman washes her clothes further downstream. Constable is famed for his desire to capture natural beauty with an eye for realism, but to the eye of the hydrologist the landscape is one modified by long-term human activity: the floodplain has been transformed into agricultural field with trees limited to the banks and hedgerows; the river itself has been co-opted for human use, its riverbanks secured against erosion underneath the footbridge; the perspective of the painting places the viewer with their back towards Flatford Mill and Lock where the river channel has been straightened to help industrial processes and ease navigation.

'View on the Stour near Dedham' by John Constable (1822). Private collection

One of the first rivers in England to be made navigable by an Act of Parliament in 1705, by the time of Constable’s painting the Stour had already undergone extensive modifications to help the transportation of minerals, crop, and London sewage destined for the fields. Paintings like View on the Stour provide a snapshot of the long history of river modification in the UK. Despite his fondness for natural beauty, when remarking on his affection for the countryside Constable conflated natural processes with human influences, describing “its luxuriant meadow flats sprinkled with flocks and herds, its well-cultivated uplands, its woods and rivers". His perspective of the countryside is one where human activities have redefined the landscape, reshaping how populations interact with natural phenomena. The story of the Stour is familiar to many river systems throughout the UK. Currently there are few rivers with flows that can be are free enough of human influence to be considered natural – less than 15% according to the UK Centre for Ecology & Hydrology. The impact of human activities has been felt across the country with only 14% of rivers considered in Good Ecological Status under the EU Water Framework Directive as of 2019.

A key outcome is an increase in the frequency and magnitude of flood events throughout the last century: alongside changes in the climate, the floodplain has become increasingly urbanised with hard surfaces preventing rainwater from being stored in the soil. Both urbanisation and agricultural practices remove deep-rooting vegetation like trees, erasing their potency for trapping overland flow. This is exacerbated by grazing cattle which compacts the soil underhoof to make it harder for water to penetrate. Together this increases the volume and speed of rainwater flowing into rivers, exaggerating both flood levels and the speed which flood waters are conveyed downstream. More agricultural land also means loose soils can be stripped away by overland flow, increasing the sediment load in rivers beyond their natural capacity whilst bringing with it chemical fertilisers which harm river life. River channels have been straightened to aid navigation but as consequence the unimpeded flow allows flood waters to rush downstream. Dredging rivers of sediment and removing channel vegetation destroys important habitats and degrades ecosystems whilst erasing another important flow control.

Aquatic vegetation are a key concern for river managers due to their role as major flow controls: plants exert a powerful influence on river flow by blocking the channel, absorbing momentum, and generating turbulence. Altogether this slows river flow and leads to an increase in local river depths. Engineers have known about the influence of river plants since the nineteenth century when in 1869 the Swiss engineers Emile Ganguillet and Wilhem Kutter determined the mean flow resistance exhibited by vegetation in various canals and streams.

Seventeen years before Ganguillet and Kutter’s pioneering study John Millais painted Ophelia, depicting the romanticised drowning of the grief-stricken noblewoman in Shakespeare’s Hamlet. At the time the realistic depiction of riverine vegetation drew ire from contemporary critics, with one bemoaning "there must be something strangely perverse in an imagination which souses Ophelia in a weedy ditch”. But this realism portrayed a botanical scene typical for many rivers, illustrating two species of aquatic plants commonly found across the UK: branched bur-reed (emergent grasses, in the painting’s bottom-left) and stream water crowfoot (submerged, trailing canopy below and parallel to Ophelia). The two species are morphologically different and yet each are known to exhibit control on river flow: Branched bur-reed’s tall rigid stems are highly resistant, meaning its influence endures during high flows when the risk of flooding is greater; the canopy of stream water crowfoot is both dense, blocking large sections of the channel, whilst its flexibility makes the plant a potent source of turbulence which dissipates the river’s kinetic energy. The role of plants also extends beyond interacting with river hydrodynamics. Throughout the last century researchers have labelled aquatic vegetation ‘bio-engineers’ due to their ability to shape river flow to such a degree that it changes a river’s geomorphology, altering flow paths and the transportation of sediment to change the shape of river channels. It’s little wonder then that river plants have been historically seen as a contributor to flooding.

'Ophelia' by Sir John Everett Millais (1851-52). Tate Britain, London.

Flooding is the most common disaster in the world, and river managers looking to reduce the risk fo river flooding have historically removed river plants rivers indiscriminately. However, this also changes the local ecosystem as different species bounce back with different growth rates to alter the local habitat and flow control. This removal is also often inappropriate – dredging is only suitable for slow moving rivers where flow is at a speed below average walking pace. In these slower streams deposited sediment cannot be removed by natural scouring, but this is not the case for the majority of the UK’s rivers. However external, often political, pressure can motivate practitioners to adopt dredging anyway.

Despite a history of wholesale plant removal, over the last couple of decades flood management strategies have increasingly used natural processes to mediate river flow. This shift in perspective has come with an appreciation for the uncertainties involved in managing such a dynamic system and where hard defences, such as flood walls and levees, become increasingly expensive to sustain and adapt. In some cases, efforts to control river systems have led to increased flood risk: both walls and levees force flood waters that would otherwise be spread across a floodplain through narrow channels that increases the speed with which the flood peak is conveyed. The consequence is populations downstream experiencing faster flowing flood waters over a shorter time period.

An alternative is provided by aquatic vegetation: rather than being seen as an obstacle, river plants have become a tool to support traditional engineered solutions. Flooding is a natural and inevitable mechanism of many rivers. Practitioners concerned with reducing the threat posed by flooding can use river plants to work with the natural system. The selective use of plants may allow river managers to choose where the river will flood, inundating fields to use as detention zones for floodwaters and reducing the threat to human life downstream. Plants also provide additional benefits, from increasing fish stock to providing aesthetic value.

Managing aquatic vegetation is a source of uncertainty for decision-makers: the relationship between vegetation abundance and flood magnitude is unknown, and this effect may be changing. Climate change is affecting both the growing periods of plant life and the intensity of hydrological extremes. It’s expected that peak vegetation biomass in UK rivers, i.e. when the river plants are at their largest and most numerous, will occur later in the year as global average temperatures rise and seasons become warmer. With intensified storm events predicted for the colder months it’s possible peak biomass may overlap with intensified rainfall events. This combination of increases in vegetative flow control and river flow might exacerbate flooding but, again, it’s unknown by how much.

Estimated changes in peak biomass for stream water crowfoot for the UK. Environment Agency (2008)

My research within the Institute for Risk & Uncertainty is looking to address these uncertainties and better aid decision-makers. Most research investigating the effects of vegetation on river flow has either been performed in laboratory flumes, neglecting other flow controls such as river channel geometry, or uses two-dimensional simulations which neglect the three-dimensional, vertical structure of turbulence within the water column. Where one is useful for understanding the 3D dynamics of plant-flow interaction and the other is useful for broad estimations of flooding, a unified approach is needed to understand how vegetation contributes as a flood control in natural rivers.

A solution is provided by computational fluid dynamics (CFD). CFD uses the physics of fluid flow to simulate real-life conditions. It works in a similar way to a wind-tunnel: imagine a vehicle, perhaps a sports car, in the middle of a wind tunnel. Experimentalists might be interested in how aerodynamic that car is and so observe how the velocity of the wind changes as the air interacts with the car. Places where the wind slows suggests the momentum of the air flow is being absorbed, whilst an increase in wind speed suggests that the air is being deflected around the body of the vehicle. They might also be interested in how much turbulence is produced, as differences between wind speed in the tunnel and that affected by car creates instability. The resulting turbulences mixes the layers of air, colliding molecules to dissipate kinetic energy and slow the air flow. For sports cars, this may be an indicator on how much the vehicle is contributing to a slow-down in the flow as air pulls on its surface, what is known as drag. This helps to inform the car designers how much drag each car experiences and how this may slow the vehicle. Now imagine that wind as water, the vehicle as vegetation in a stream, replace the tunnel with the complex shape of a river channel, make it all virtual and you’ve got a CFD approach to simulating vegetation-flow interactions in a natural river. Altering the abundance of biomass and the magnitude of flow in line with climate change projections in turn allows researchers to inform strategists of appropriate flood management design.

Birds-eye view of a vegetated river meander during a flood event: dark spots in the main channel indicate a vegetation patch which has absorbed the flow's momentum, resulting in wakes or 'shadows' extending behind each patch where flow speed is less than the surrounding water.

Interest in natural solutions to flooding comes at a time when 90% of the UK’s floodplains no longer function properly: their ability to store floodwaters is in decline due to human alterations whilst becoming more populated. Better management of vegetation, both on and off the floodplain, can help regain some of the natural mechanisms that regulate flooding otherwise lost to the centuries of human interference. Understanding the magnitude to which these mechanisms impact flood events and to what extent these will be altered by future climate change will provide a key insight to decision-makers and river managers looking to reduce the impacts of flooding.