A parameterization of tidal marsh inundation was developed for the 1,200 hectares of tidal marsh along the 12-km reach of the tidal Murderkill River between Frederica and Bowers Beach in Kent County, Delaware. A parsimonious modeling approach was used that bridges the gap between the simple and often used “bathtub model” (instantaneous inundation based on tides in Delaware Bay), and the more complex modeling of shallow overland that results in the wetting and drying of tidal marshes. For this project, and many other modeling studies that include large areas of marsh, a complex modeling approach of marsh inundation is not warranted due to the lack of data on the dynamics of wetting and drying. A simple parameterization of the wetland inundation process coupled with more complex hydrodynamic and water-quality models can provide sufficient results for estimating the extent of hydrologic and biogeochemical interactions between a marsh and a river. The parameterization can also be used to evaluate anomalies in conservation of water mass and tidal phase offsets that can result from hydrodynamic models that do not explicitly model the dynamic flow and storage of water in tidal wetlands. In the parameterization, the marsh was divided into marsh tracts (n=31) based on hydrologic character and position along the river. A cumulative probability distribution of wetland elevation was calculated for each marsh tract from a digital elevation model. These cumulative probability distributions served as a simplification of the critical information contained in the raster data sets of marsh tracts and elevation. Each marsh tract was related to an adjacent river reach; the area in the tract that was below the stage of its related river reach was instantaneously inundated. Marsh tracts were aggregated into two sets of marsh groups (n=22 and n=4) for analysis and visualization of elevation, hydroperiod, and hydraulic loading. The parameterization was successfully implemented in a collaborative modeling study that created a set of mass loading functions to represent the import and export of chemical species to and from the wetlands. The parameterization was also used to evaluate conservation of water mass and phase offsets in tidal discharge due to the dynamic storage of water in intertidal areas.
Marsh elevations had a normal distribution with a mean elevation of 0.72 m and standard deviation of 0.19 m based on analysis of LiDAR data collected for this study. These values have a potential positive bias of 0.1 to 0.2 m resulting from the LiDAR beam not penetrating through the marsh vegetation. Nominal relief on the marsh at the scale of the study area was about 0.6 m (0.4 to 1 m absolute elevation using the NAVD88 datum). From Bowers Beach upstream to Frederica there was a decrease in marsh elevation with the mean elevation decreasing from 0.86 m to 0.60 m. This observation is consistent with measured accretion rates at four sites in the study area that document higher accretion rates upstream near Frederica (0.74 cm/yr) relative to downstream near Bowers (0.33 cm/yr). Upstream marshes are flooded more frequently and for longer duration than downstream marshes so there is more opportunity for accretion to occur.
University of Delaware
Delaware Geological Survey Building
Newark, DE 19716
Mon - Fri; 8:00am to 4:30pm