When can rewetting of forested peatlands reduce extreme flows?

Historical drainage to improve agricultural and forestry practices has resulted in almost 1 million km of artificial channels in Sweden. This has reduced the storage of water in the landscape, and there are concerns related to the potential impacts on extreme flows, biodiversity, greenhouse gas emissions and nutrient outputs. A large national restoration program aims to rewet 100000 hectares forested peatland. However, there is limited evidence in what the impacts will be. Here, we implemented national information on ditches to the hydrological model HYPE and investigated the conditions at which removal of ditches in forested peatland could mitigate extreme flows under various conditions of the climate and local hydrology. We found that the impact on discharge at the level of 10 km2 sub-catchments or larger was small, mostly because only small fractions of the catchments consist of drained forested peatlands, meaning there is considerable mixing with other runoff. However, smaller streams with runoff primarily from the restored peatlands could have substantial impacts of restoration, which may be important for local biodiversity. For instance, a modelling sensitivity study showed the minimum runoff per year from forested peatlands increased by up to about 15 % after removing ditches and the maximum runoff was reduced by up to about 25 %. Importantly, an increase in the minimum runoff was only obtained if the minimum groundwater level was low enough in relation to the depth of ditches. Similarly, a reduction of the largest yearly runoff required that ditches were not too deep. If conditions were not favorable to mitigate extreme runoff, the opposite situation often occurred instead, with worse extremes. Therefore, although the impact on extreme flows was negligible at the level of 10 km2 catchments or larger, it is crucial to choose appropriate sites for restoration with respect to runoff extremes if there are sensitive smaller streams with runoff deriving mostly from the peatlands. The work presented here shows how this can be performed with the use of indicators for groundwater levels and ditch drainage prior to rewetting. Specifically, the minimum runoff is expected to increase only if the minimum groundwater level prior to rewetting is below the depth of ditches, or close to that depth. Reductions in the maximum runoff require ditches are not too deep, and large reductions cannot be expected if the groundwater level was already temporarily above the soil surface prior to rewetting, for example due to lateral inflow.