Flood is a natural peril but globally there is a tendency to manage flood risk with man-made measures. More and more people live in waterfront locations, either on riverbanks or on coasts, and the typical methods for trying to tame rivers or the ocean involve massive engineering projects and expensive flood defenses. These are often rigid and inflexible, and while effective to high design specifications, if they fail, catastrophe ensues. More flexible and resilient approaches can be used in conjunction and in place of these hard physical solutions, and often involve nature-based solutions which can allow a more staged reduction of flood risk and extra benefits to habitats and ecosystems.
Natural flood management (NFM) refers to the use of natural processes and environments to mitigate flood risk1. NFM is increasingly being viewed as a cost-effective supplement to traditional flood defense methods, which require the construction of hard infrastructure such as dams, weirs, and levees.
Some NFM techniques, as classified by the Environment Agency in England, include river restoration, floodplain restoration, leaky barriers, offline storage areas, woodland creation and conservation, soil and land management, headwater management, and runoff management2. Although a holistic approach to flood management is likely to include several, if not all, of the above techniques, research on NFM is still in its early stages and we still do not know enough about the effectiveness of various interventions.
In a recently published article in the WIREs Water journal3, the current state of knowledge on the use of forested lands for NFM in the UK was reviewed. The review specifically focused on four woodland types: catchment, cross-slope, floodplain, and riparian. The key finding was that, overall, we are still very limited in the amount of published evidence available on the ability of forested terrain to mitigate flood risk, and while flood impact continue to be a major topic of conversations in the risk management industry, more targeted research is needed to improve uptake and efficacy of these approaches.
Evidence on the impact of different woodland types on peak flow reduction in rivers is highly variable. For catchment woodland, there is a rich history of paired catchment studies in the UK, which involve one catchment being kept as a control (forested) and a nearby similar catchment being treated (i.e., thinned or clear-felled). However, the findings from these studies have been mixed.
At the Coalburn catchment in north England, which is Britain's longest running forest hydrology research catchment, the first five years after clear-cut showed that peak flows increased by 20% and the time to peak decreased by about a third; while 10 years later the peak flows were still 10% above the pre clear-cut values4. In contrast, studies at the Plynlimon catchment in mid Wales and the Balquhidder catchment in mid-Scotland did not show any significant changes to peak flows after clear felling of forested areas5, 6.
For cross-slope woodland, most of the evidence for peak flow reduction benefits exists at farm or hillslope scales. Several studies have shown that planting cross-slope woodland provides a way of reducing the impact of livestock on soil infiltration rates and can help reduce the risk of rapid runoff from farms7. A noteworthy study here is one by the Flood Risk Management Research Consortium at the Nant Pontbren catchment in Wales8, which showed that optimally placed cross-slope woodland shelter belts could potentially reduce peak flows for frequent events by 29%, compared to a 50% reduction for complete tree cover across the catchment.
There are few empirical studies investigating the impact of floodplain and riparian woodland on flood peaks and, unfortunately, most of the evidence consists of limited scenario-based modelling exercises conducted with hydraulic models such as HEC-RAS, River2D, and InfoWorks RS9. Nonetheless, results from the modelling studies encouragingly show that, at smaller scales at least, restoration of floodplain and riparian woodland would decrease the flow rate of water within the planted area, raise the water level, delay peak discharges, and desynchronize the flood peaks from adjacent tributaries.
Despite the variability of available evidence, all woodland types need more research to effectively guide the future development of forest planting proposals, and to enable the most effective use of forested lands in contributing to NFM. Further improvements are also required in modelling before a generalized conceptual model is available that accounts for variation in controlling factors at the stand, hillslope, and catchment level. However, these efforts will need to be guided by an increase in observational evidence through field and remotely sensed measurements.
It will take time, but these are research areas worth pursuing. Collective knowledge on flood mitigation impacts of NFM techniques will help us advise governments better and justify the necessary investment in such natural solutions.
1 https://doi.org/10.1002/wat2.1211.
2 Environment Agency. (2018). Working with natural processes – Evidence directory. Environment Agency, 311.
3 https://doi.org/10.1002/wat2.1541.
4 https://doi.org/10.1016/0022-1694(86)90007-7.
5 https://doi.org/10.1002/hyp.5561.
6 Johnson, R. C. (1995). Effects of upland afforestation on water resources – The Balquhidder Experiment 1981–1991. Institute of Hydrology.
7 https://doi.org/10.1002/hyp.10630.
9 https://doi.org/10.1111/j.1747-6593.2006.00056.x.
