Taking an integrated landscape approach to Natural Flood Management

By Prof Chris J Spray MBE, FCIEEM, Tweed Forum Eddleston Water Science Manager

Drone image of a new pond surrounded by grasses and trees — A newly created pond at Cringeltie (Tweed Forum)

Introduction
Agnes, Babet, Ciarán and now Debi….this Autumn’s storms have hit the headlines and flooding has rarely been higher on the UK agenda (and globally). With climate change disrupting established patterns and intensity of rainfall events, made worse by land use change, drainage and building in the (increasingly disconnected) floodplain, focus has turned to the potential for nature-based solutions (NbS) to work alongside ‘traditional’ flood defence schemes to help reduce flood risk. Indeed, the Scottish 2009 Flood Risk Management Act sets a legal requirement for considering what the ‘natural characteristics’ of a catchment could offer in terms of reducing flood risk downstream. To do so requires looking at the whole catchment landscape, recognising that floods have sources, pathways and receptors. But this also raises questions as to applicability and effectiveness of natural flood management (NFM) to reduce flood risk and enhance riparian habitats.

The Eddleston Water project explores this in a typical 69km2 catchment in the Scottish Borders, where the risk of flooding to Peebles downstream is real and immediate. The challenge Scottish Government (main funders) set and which Tweed Forum (project managers) along with partners SEPA and Scottish Borders Council accepted was simple: could one provide the evidence that NFM works so as to underpin policy and practice in this area? Working with University of Dundee who produced the scoping study in 2010 and British Geological Survey, the study is now in its 13th year of ‘learning by doing’. Evidence gathering covers both biophysical and behavioural perspectives - hydrology, hydromorphology, ecology, environmental economics and the social sciences.

Newly re-meandered river with bare banks and new fencing — The freshly cut re-meander at Lake Wood (Sept 2013) (Tweed Forum)

Options for landscape-scale action:
Tackling the generation of floods at source: Uplands, as the gathering grounds offer opportunities for NFM interventions to reduce the generation of floods in the first place.

Woodland planting helps deliver reductions in flood risk: -through intercepting rainfall in the canopy (UK annual losses: 32-45% conifers, 17-23% broadleaves); by increasing soil infiltration through their root systems; and by increasing hydraulic roughness, acting as a barrier to surface water flow - together delaying flood peaks and enhancing floodplain storage. So far, we have planted 210ha of native trees across Eddleston and scaling this up through modelling landscape-scale tree planting under different climate change scenarios shows up to 40% reduction in peak flows, and flood peaks delayed by 45 minutes. Alongside this our empirical measurements of infiltration under mature broadleaf woodland show it is 5-6 x that under adjacent grazed sheep pastures and pine plantations (due to the former’s extensive, deep roots), thus reducing rapid overland water flow.

High-flow restrictor log structures are an effective way to delay and dampen flood peaks and we have installed 116 of these leaky barriers. Those upstream in the Middle burn produce a significant increase in ‘lag time’ between rainfall arriving and rising river levels downstream of 2.6-7.3 hours. Taken with other NFM measures, the flood peak in the upper catchment has been reduced by c 30% following their deployment. These are different to beaver dams as they permit low flows to pass unhindered and so do not create a ponded landscape upstream.

Logs grouped together and laying above the water across a small creek — A simple high-flow restrictor log structure (a leaky dam) on the Middle burn (Sept 2023) (Tweed Forum)

Ponds can be designed to be temporary flood storage areas. We have created 36 flow attenuation ponds located in the headwaters, and 2 large ones on the lower floodplain. Pond storage volume is generally small compared with catchment area and associated runoff, but there is potential for such ponds to help control flood peaks for small flood events, as long as they are carefully designed (engineering of inflow/outflow structures) and sufficient in number.

Pathway opportunities: the mid catchment and floodplain of river systems offer additional opportunities, notably reconnection of embanked rivers to their floodplains, opening up large areas of temporary storage and the remeandering of previously-straightened watercourses. With one exception at Kidston Mill, we have not developed large floodplain ponds in the Eddleston, because the floodplain is productive and valuable agricultural land, the flooding of which is economically unviable in the current system. That said, several such ponds together could be very effective in low intensity events, providing large areas of temporary storage, which at a local level may be able to reduce the discharge peak by some 18–20% and theoretically delay it by up to six hours.

Remeandering of the channel (straightened 200 years ago) has been undertaken over 3.5km in various sections. Where this has also included removal of flood banks this has led to up to 6% additional flood storage, but mostly this is less and remeandering alone produces small flood risk benefits.

Receptors are the properties at risk of flooding and by the time a flood pulse reaches here, the focus largely turns to traditional forms of flood defence. This, along with better flood warning schemes and controls on building in the floodplain should be seen as complementary to, not opposed to NFM. And whilst they have not been the focus of the Eddleston study, there are still opportunities for sustainable urban drainage measures.

Huge flooding throughout a town with a bridge just visible above the water — Flooding downstream in Peebles (Tweed Forum)

The added Biodiversity Bonus of NBS:
One of the attractions of an NFM approach is that measures can also deliver significant improvements for biodiversity. Whilst high-flow log structures do not deliver the biodiversity and water quality benefits associated with beaver dams, other NFM measures deliver significant improvements to habitats and wildlife.

Riparian trees used as an NFM measure are good for biodiversity but also provide shade which is reflected in lower water temperatures recorded at sites along the Eddleston of up to 1.5°C; important to salmonids needing to adapt to rising temperatures with climate change.

Ponds can be designed so as to always hold some water, whilst having the capacity to accept much larger volumes in flood periods. Our work on Eddleston looking at macroinvertebrates and especially dragonflies in the ponds created for flood storage shows that they are as good, if not better than similar UK ponds designed for wildlife conservation in terms of species richness.

Remeandering straightened channels not only results in an increase in channel length (8-46% in remeandered reaches), which in itself leads to more available habitat, but once recovered from disturbance, these habitats show increased diversity compared with previously straightened channels.

Conclusions:
NFM will not of its own counter rising flood levels, and neither will it deliver the step-changes in wetland and wildlife restoration needed, but as an NbS it has a role to play. Evidence for its success is best for impacting small floods and in small catchments (<10km2), but there is increasing evidence being produced for its ability to help reduce risk in larger floods and larger catchments. What is clear though is that it will need a diverse range of measures across the whole landscape and that if carefully designed and placed these measures can also deliver significant gains for biodiversity.

Chris can be contacted on: C.J.Spray@dundee.ac.uk
Website for the project is here: https://tweedforum.org/our-work/projects/the-eddleston-water-project/