CLEARANCE_Potential for wetland restoration in ...

Wetlands have the capacity to remove nitrogen from surface waters in the process of denitrification. Odense River catchment has been a subject of intensive wetland restoration to mitigate N loss and restored wetland area has been reported to reduce the N load to Odense river with 124 tons N per year. The objective of this work was that further wetland restoration in Odense River catchment would result in further reductions in N load to coastal waters. The potential wetland area was found by creating an index model that showed suitable areas. Direct upland area to each wetland was calculated in ArcGIS, N loss from each upland was calculated and removal from the direct upland was estimated based on soil type and drainage probability. N removal by flood inundation was based on a flood estimate accounting for 10% of the time, which was calculated based on measured data of stream water level increase from monitoring stations and the upstream river length (Upstream Length Model, UPM). The flood based on the UPM provided a very good flood estimate, that evenly distributed flood upstream as well as downstream and required minimum data inputs. Compared to flood based on MIKE11 calculations of water level increase, the UPM performed better, as MIKE11 resulted in a very unlikely overestimation in one of the streams, which shouldn’t be the case considering the amount of input data used to calculate water level elevations in MIKE11. The UPM can be further improved during stream simplification, which is a necessary step to provide a satisfying interpolation of elevation raster used in flood calculation in ArcGIS.
The N load and removal in already restored wetlands was also calculated and amounted to 75,094 kg N. The comparison between calculated N removal against measured data from to restored wetlands showed the calculated values to be underestimated. That is due to the flood calculation accounting for only 36 days, while many restored wetlands are designed to be flooded for longer time periods. Another issue is that N removal in shallow lakes couldn’t be calculated due to lack of data on lake residency time.
The restoration of an additional 5543.9 ha of wetlands was calculated to remove 405,392 kg N. That calculation didn’t include N removal in shallow lakes, so the potential N removal could be higher. The highest N removal per ha was found in wetlands irrigated with drainage water and was caused by a higher percentage of agricultural cropping in the direct upland area compared to non-drained upland area. Several irrigated wetlands had very high calculated N loads, where an expected removal rate of 50% is questionable, as according to measured data high N loads are associated with decreased removal rates. The actual N removal in wetlands with non-drained upland area could be higher than calculated, as many of them could be groundwater fed. The calculated flood N removal could be more accurate in potential wetlands than restored wetlands, as many streams overflood for approximately a month yearly. As N loss varies between the direct upland areas, the potential wetland area can be modified to yield higher N removal per wetland area.