Setup of hydro-morphodynamic model

To understand the effects of different flow regimes, seasonality and regulation of flow into a re-opened polder section on sedimentation inside a beel with TRM, we used a two-dimensional numerical hydro-morphodynamic model that was developed and calibrated for Beel Pakhimara, which is an active TRM in southwest Bangladesh (Islam et al. 2019). Model simulations were carried out using Mike 21FM (developed by DHI) in which hydrodynamic processes and sediment transport are simulated simultaneously (DHI, 2012a). As the grain size of the sediment is fine (less than 63 µm) for the study area (Datta & Subramanian, 1997; IWM, 2010), the MT (mud transport) module of Mike 21FM was used for calculating cohesive sediment transport.
The hydrodynamic (HD) processes of Mike 21FM are calculated based on the solution of three-dimensional incompressible Reynolds averaged Navier-Stokes equations (DHI, 2012a). The model uses an approximate Riemann solver to calculate the convective fluxes at the interface of the cell of the 2D mesh (DHI, 2012a). The MT module uses the advection-dispersion equation (ADE). The ADE is solved using the third-order finite difference scheme, known as the ULTIMATE scheme, which is based on the QUICKEST scheme (DHI, 2012b). For morphological simulation, the bathymetry is updated for each time step according to net sedimentation (DHI, 2012b).
To represent the bathymetry of Beel Pakhimara, a mesh with flexible cell size was used where the 2D cells had shapes from triangle to octagon. Finer mesh was opted to represent the inlet canal (cell area of about 170 m2) and coarser mesh (cell area up to 5000 m2) was used for the flood plain to reduce the calculation intensity (Islam et al., 2019). For our analysis we supplied the model for Beel Pakhimara with water levels and SSC, that were adapted in accordance to the scenarios of different flow regime and seasonality (Figure 2, 4 and 5). These input variables were provided for Locations 1, 2 and 3 with respect to their flow regime (Figure 1).
The model was calibrated for Beel Pakhimara by comparing the observed water level, discharge and SSC with simulated ones (Islam et al., 2019). Manning’s coefficient, shear stress and settling velocity were the primary parameters for calibrating hydrodynamic and morphodynamic model. Sensitivity analysis of the model was carried out with varying Manning’s coefficient from 0.1 s/m1/3 to 0.01 s/m1/3, shear stress from 0.01 N/m2 to 0.1 N/m2 and settling velocity from 0.0001 m/s to 0.001 m/s. To understand the uncertainty of the model, coefficient of determination (R2) and the normalized root mean square error (NRMSE) were calculated by comparing the modelled results with the observed data for the different input variables. For spatial average value of Manning’s coefficient of 0.032 s/m1/3, shear stress of 0.08 N/m2 and settling velocity of 0.0005 m/s, the R2 for water level, discharge and sediment concentration were 0.87, 0.88 and 0.84, respectively. The NRMSE (%) for water level, discharge and sediment concentration were 9.7, 16.6 and 18.3, respectively (Islam et al., 2019). The calculated goodness of fit for the developed model indicates that the calibrated model resulted in good agreement between the observed and simulated data.

Scenario development

The calibrated model for Beel Pakhimara was used to simulate the developed scenarios for different regions. Time variant water level and SSC representing different flow regimes were used as input variables to investigate the sediment dynamics of different regions. To compensate for the difference in land level between the river-dominated and the tide-dominated region, the water levels used for location 1 representing river-dominated region were vertically shifted accordingly. The mean difference between the water level at location 1 and location 3 (Figure 1) during dry season was calculated as 1.03 m. The water level of location 1 was lowered by 1.03 m to compensate for the land level difference between river-dominated and tide-dominated region. Sediment rating curves presented in IWM (2017) were used to generate input variables for the model when measured SSC data were unavailable. As data was not available continuously in all three locations, model simulations were carried out for 14 days consecutive to capture the effect of spring and neap tide for each season and to have consistent scenarios for all three locations. It was assumed that the simulated periods are representative of the seasons.
The scenarios were defined by varying flow regimes, seasonality and flow regulation schemes into the beel within the polder. In addition to unregulated flow into the beel, gate operation rules for the regulated flow scenarios were adopted from Islam et al. (2019): one or both gates remained open for 12 hours to allow the flow of sediment-rich water to enter the beel and were closed for the following 12 hours. By doing so, no flow inside the beel occurs for 12 hours allowing the sediment to settle from the stagnant water. Two different rules of gate operation were considered: simultaneous and successive (Islam et al., 2019).  For simultaneous gate operation, both gates are opened and closed at the same time for both inlets (Islam et al., 2019). For the successive gate operation of two inlets, their gates are alternatingly open and closed (Islam et al., 2019). In that case the flow enters through one inlet, and leaves the beel through the other. This gate operation aims to ensure throughflow across the beel instead filling and draining through the same gate. The gates in Bangladesh are manually operated, and operation is time consuming. Therefore, it was assumed that the opening and closing of the gate will require 1 hour. The developed scenarios by combining flow regimes, seasonality and flow regulations are presented in Table 1.