Flood mitigation at catchment scale: assessing the effectiveness of constructed wetlands

Abstract: Considering the potential for natural wetlands to store and retain water in the landscape,restored and constructed wetlands are increasingly implemented as Nature‐BasedSolutions (NBS) to mitigate downstream flooding. However, there is a frequent lack ofknowledge of the performance of NBS, and particularly wetlands, and how to best selectperformance indicators to measure their effectiveness. The aim of this study was to assessthe flood regulation services of multiple constructed wetlands at catchment scale. TheSWAT+ model was used to assess the impacts of 35 constructed wetlands in the 193 km2Råån catchment in southwest Sweden. The assessment was based on a paired simulationscenario, with and without wetlands, and performance was evaluated based on impacts onthe downstream river hydrograph (peak flow rate, time-to-peak, recession curve shape,daily streamflow average) as well as event-based (daily average streamflow for differentprecipitation ranges), descriptive (average wetland storage capacity), consequential(overbank flooding) and social (social acceptance) indicators. The results implied that theflow regulating capability of the modeled wetlands was directly related to their outflowmechanisms. The wetlands had an average storage capacity of 0.38 m3 per m2 area butbecause the wetlands seemed to frequently exceed their maximum storage capacity duringhigh flow events, and the excess water was immediately spilled downstream, the wetlandsprovided limited flood regulation services. There were no visible impacts on thehydrograph-related indicators and the daily streamflow average was reduced by anegligible 0.02%. Furthermore, there were no significant differences in wetlandperformance between small and heavy rainfall events and the model output showed nosigns of overbank flooding for either wetland scenario. In terms of social acceptance, recentwetland projects in Råån catchment seem to have been well-received by farmers andlandowners in the area. However, lack of wetland-specific data, such as volume and dailyinflow and outflow data, as well as other model uncertainties, motivated a low confidencegrade for the indicators. In fact, if indicator values are to be useful tools in the futureintegration of wetlands as nature-based solutions, in-situ monitoring data of wetland waterbalance components are required. To support future policies and decision-making, adaptedtools and further research are needed to assess the impact of NBS on upstream-downstream processes.