The potential of woodland to prevent flooding has been enthusiastically promoted in recent years, particularly by environmental interest groups such as the Woodland Trust and the WWF (see Woodland Trust, 2012, p13-15 and WWF Scotland, 2007, p2). However, some commentators have questioned the robustness of the link between woodland cover and flooding (see Calder & Aylward, 2006). How effective is woodland at reducing flood risk? Can afforestation and woodland management play a significant role in reducing flood risk, or is the effect of trees minor when compared to other flood risk management interventions?
One mechanism by which woodland could affect catchment hydrology, and therefore flooding, is in how the trees affect water transfer into the soil. Root growth through soils can lead to increased soil porosity, enhancing infiltration and therefore reducing overland flow. Some research has found that infiltration rates are up to 60 times higher in stands of new trees than in nearby grazed pasture (Carroll et al, 2004; Marshall et al, 2009), which lends support to the theory that trees can encourage infiltration and reduce overland flow – at least at plot- and field-scale. However, other larger-scale studies found no link between deforestation and flooding (see Crook, 2002, for a study of flooding and forest cover in a 150 km2 catchment over a 250 year period, based on pollen records).
Another potential mechanism by which woodland could affect hillslope hydrology is by physically altering the overland flows of water. Interception or interruption by vegetation increases the friction experienced by flows and increases the distance travelled. This could therefore hold water back within woodland and attenuate flood peaks. Similarly, evapotranspiration could also affect the volume and speed of water reaching the channel. The ongoing uptake of soil water by roots, and its transportation to the leaves, can lead to reduced soil moisture levels in times of low or normal rainfall. This could potentially give increased water storage capacity during storms. Other more indirect mechanisms by which woodland could affect flood risk include the stabilisation of river banks and the generation of large woody debris (LWD).
It therefore seems conceptually attractive that woodland could reduce the volume and speed of water reaching the channel, and therefore decrease flood peaks downstream. At a catchment scale, research has found that in some places there is “considerable scope for using strategically placed floodplain woodland to alleviate downstream flooding” (Thomas & Nisbet, 2007). This study used a hydraulic model to investigate the effect of riparian afforestation on flood flows, considering the performance of a reach with and without broadleaf woodland along a section of riverbank. The results showed that water velocity across the floodplain was substantially reduced, increasing floodplain storage by up to 71% and extending the flood peak travel time by 140 minutes. Crucially, in this case the effects were seen with only a relatively small area of the catchment being afforested (less than 2% of the total 82.4 km2 catchment).
While this study produced encouraging results, it also showed that woodland has a complex relationship with the catchment in which it is sited. This is highlighted by the authors, who emphasise the ‘strategic’ placement of floodplain woodland. In other words the siting of woodland is crucial, rather than simply its presence or absence in a catchment. Other research has suggested that woodland’s role in attenuating flows in larger catchments and during more extreme flood events is likely to be rather more limited (Nisbet, 2002). Nisbet also identifies that the soil cultivation and drainage associated with commercial forestry can have a substantial impact in increasing the volume and speed of flood flows, so afforestation schemes should be carefully planned and managed to minimise any adverse effects.
When considering woodland creation or management, it seems therefore that the emphasis should be on strategic placement of woodland within a catchment, with appropriately rigorous modelling. There is likely to be little gained, from a flood risk perspective, from indiscriminate or piecemeal afforestation (although there may of course be significant benefits in terms of amenity, carbon sequestration or biodiversity).
The most persuasive examples seem to come from modelling of inundated wooded floodplains. As well as Thomas & Nisbet’s work, Dixon (2014) found ‘substantial and predictable responses in downstream flood height’ from modelling a combination of LWD and upstream floodplain afforestation. Of course, accessing the benefits of floodplain storage first requires floodplain inundation, and the associated complexities of dealing with floodplain landowners and stakeholders (for examples, see Howgate and Kenyon, 2009).
When considering woodland creation or management the emphasis should therefore be on:
- Strategic placement of woodland within catchment, with appropriately rigorous modelling
- Ensuring intervention is of appropriate scale
- Minimising any adverse effects associated with commercial forestry
- Maximisation of other short-term benefits due to long pay-back period
This article originally appeared in the Blue-Green Cities weblog.