RIV: River Package Pane

The RIV: River package pane is on the MODFLOW Features tab of the Object Properties dialog box.

The data that can be specified for the River package are the Starting time, Ending time, River stage, Conductance, and River bottom. The River stage is the head in the river. River bottom is the elevation of the bottom of the river bed.

Water can leave the groundwater through the river boundary when the head in the cell is higher than the River stage. The rate of flow will be proportional to the difference between the River stage and the head in the cell.

Water can enter the groundwater at a variable rate through the river boundary when the head in the cell is below than the River stage but higher than the River bottom. The rate of flow will be proportional to the difference between the River stage and the head in the cell.

Water can enter the groundwater at a constant rate through the river boundary when the head in the cell is below than the River bottom. The rate of flow will be proportional to the difference between the River stage and the River bottom.

The Conductance is the factor that relates the difference in head to the rate of flow.

Starting time and Ending time are explained in the help for the MODFLOW Features tab.

If parameters are used with an object, River stage, and River bottom must be defined for each parameter as well as a multiplier for the Conductance. MODFLOW will multiply the multiplier by the parameter value (specified in the MODFLOW Packages and Programs dialog box) to determine the conductance to apply.

In MODFLOW, the Conductance is described as having units of L2/T and is equal to KLW/M where

K = the hydraulic conductivity of the sediment in the boundary condition such as a river or drain,

L = the length of the boundary condition in the cell,

W = the width of the boundary condition, and

M = the thickness of the sediment in the boundary condition perpendicular to flow between the boundary and the cell. Usually this will be vertical thickness of the sediment.

However, the conductance is also affected by the grid cell size independent of the above formula (Mehl and Hill, 2010).

For Polyline Objects, the length of the object in the cell can be determined with the function ObjectSectionIntersectLength.

One way to to set the conductance would be with a formula that includes ObjectSectionIntersectLength and user defined Global Variables for the width of the drain and the thickness of the sediment in the drain and perhaps the hydraulic conductivity of the sediment in the drain. For example, the user could define Global Variables named "DrainWidth" and "DrainSedimentThickness" and then set the Conductance of the drain using the following formula.  

((Kz * ObjectSectionIntersectLength) * DrainWidth) / DrainSedimentThickness

In the formula, it is assumed that the hydraulic conductivity of the sediment in the drain is the same of the vertical hydraulic conductivity of the cell that contains the drain. This may or may not be a valid assumption in any particular case.

A similar formula appropriate for Polygon Objects would be

(Kz * ObjectIntersectArea) / DrainSedimentThickness

However, if Conductance Interpretation is set to Calculated, as it is by default, ObjectSectionIntersectLength and ObjectIntersectArea should not be included in the formulas because they will be incorporated automatically for polyline and polygon objects. Thus, if Conductance Interpretation is set to Calculated for a Polyline Object, Conductance should be treated as if it were KW/M instead of KLW/M, its units would be L/t, and an appropriate formula would be

(Kz * DrainWidth) / DrainSedimentThickness

Conductance Interpretation has no effect for point objects. For point objects, the formulas will be applied without any modification regardless of Conductance Interpretation.

if Conductance Interpretation is set to Calculated for a Polygon Object, Conductance should be treated as if it were K/M instead of KLW/M, its units would be 1/t, and an appropriate formula would be

Kz / DrainSedimentThickness

If Conductance Interpretation is set to Direct, the formulas will be applied without any modification.

If Conductance Interpretation is set to Total the formulas will be modified by multiplying the conductance by ObjectSectionIntersectLength / ObjectLength or ObjectIntersectArea / ObjectArea.

Setting the Conductance Interpretation to Calculated or Total allows ModelMuse to calculate an appropriate conductance for a cell based on the length or area of intersection between the object and cell. However, if you calibrate your model and then change the grid size, the conductances assigned by ModelMuse may no longer be appropriate and you may have to recalibrate your model. The conductance is a coefficient of proportionality between the head difference in the cell and the flow to the boundary. When the grid size changes, the head in the cell will be averaged over a different area than before. This conductances will need to be adjusted because of this change in head.

Time-series interpolation

In MODFLOW 6, time series files can specify the times and values to use for a boundary condition.

When MODFLOW 6 needs data from a time series or time-array series for a time interval representing a time step or subtime step, the series is queried to provide a time-averaged value or array of values for the requested time interval. For each series, the user specifies an interpolation method that determines how the value is assumed to behave between listed times. The interpolation method thus determines how the time averaging is performed. When a time-array series is used, interpolation is performed on an element-by-element basis to generate a 2-D array of interpolated values as needed.

The supported interpolation methods are STEPWISE, LINEAR, and LINEAREND. When the STEPWISE interpolation method is used, the value is assumed to remain constant at the value specified in one time-series record until the time listed in the subsequent record, when the value changes abruptly to the new value. In the LINEAR interpolation method, the value is assumed to change linearly between times listed in sequential records. LINEAREND is like LINEAR, except that instead of using the average value over a time step, the value at the end of a time step is used.

ModelMuse uses time-series interpolation when parameters are used. The parameter value becomes the SFAC (scale factor) in the time-series file.

ModelMuse sets the interpolation method to STEPWISE for all boundary conditions in which it is used except for the CHD package in which it is set to LINEAREND.

ModelMuse does not currently support time-series interpolation because it slows down MODFLOW to much as implemented in ModelMuse. Support for time-series interpolation may be supported in a future version of ModelMuse but in a different form.