# Boundary Condition - River - Delft3D

## Modelling Rivers

## RiverRivers are natural water courses. They have vital contributions to the social and economical welfare of societies living along their course. They provide continuous, usually dependable, flow and supply of water. Accordingly, centralised civilisation began emerging independently along number of river valleys; e.g. the Nile, Tigris and Euphrates, the Indus and the Yellow Rivers. Since then, mankind has been in continuous struggle to understand and live in harmony with this gift/force of nature. Modelling rivers is the modern way to understand their behaviour and is an essential tool to tackle several river problems. Delft3D is the state-of-the-art modelling framework for hydrodynamics, water quality, ecology, waves and morphology. Here, in this space, we share knowledge & experiences, and discuss issues related to modelling of rivers.
| |

## Message Boards

### Boundary Condition

##### Ben Müller, modified 2 Years ago.

#### Boundary Condition

Youngling Posts: 2 Join Date: 10/20/17 Recent PostsI want to build a model of a river. The upper boundary condition is to be defined as a total discharge. The energy gradient gradient of 0.01% is known for the lower boundary condition. My problem is which type of open boundary is to be used for the lower boundary condition?

Thank you,

Ben Müller

##### Richard Measures, modified 2 Years ago.

#### RE: Boundary Condition

Jedi Knight Posts: 178 Join Date: 3/23/11 Recent PostsWhen I am building river models there are three options I consider for downstream boundaries, each with different strengths and weaknesses:

Option 1: Time varying (or constant) water level boundary

Option 2: Water level vs flow boundary

Option 3: Neumann (water level gradient) boundary

Option 1 gives you complete control of the boundary condition and is relevant if you have reliable data on downstream water level (e.g. there is a water level recorder at your downstream boundary or your river ends at a lake or the sea. It is difficult to apply if you don;t know in advance what the water level should be, and it cannot cope well with water level variation across the length of the boundary (i.e. across the width of the river).

Option 2 can be specified by doing a 1D normal depth calculation to generate a rating curve outside of delft3d, or by taking a rating curve from some other source (e.g. if you're boundary is at a flow gauging station with a known water level vs flow relationship). The advantage of this option is that it can cope well with time varying flows as the boundary responds to the change in flow. Disadvantages are that it can't deal with temporally varying backwater effects (e.g. in a tidal reach) and (like option1) it can't cope with water level variation across the width of the river. For stability purposes I recommend extending the rating curve well beyond the expected flow range in case there are any local flow variations (e.g. waves during model spin up from initial conditions).

Option 3 only requires you to specify a slope and can be a good solution. It has the advantages that it allows flow variability across the width of the river (important in a braided river or similar) and it allows the model to determine the water level (good if you don;t have any other data). It is important to remember that the slope you specify is the

__local__slope - so is strongly influenced by local bed roughness and bathymetry. This means you need to be very careful with roughness and bathymetry at your boundary and if you make changes (for example to roughness during calibration) then this can have a big effect on boundary water level. In the past I have found this boundary to be less stable for some rivers.

I have used all three options with success in different models, it is just a matter of selecting the most appropriate for your example.

Hope that helps,

Richard

##### Ben Müller, modified 2 Years ago.

#### RE: Boundary Condition

Youngling Posts: 2 Join Date: 10/20/17 Recent PostsThanks a lot for your answer.

Options 1 and 2 are not suitable for my model. I think I have to use option 3 but I do not know which numerical value is to be determined. I read in the manual the value of the Neumann Boundary has to be between 1 and 100 [-] . The energy gradient of the river was determined to be 0.01%. What does this mean for the Neumann boundary condition?

Furthermore, the river splits into 3 outflows so that 3 different Neumann boundary conditions must be specified.

Thank you for your help.

Ben

##### Richard Measures, modified 2 Years ago.

#### RE: Boundary Condition

Jedi Knight Posts: 178 Join Date: 3/23/11 Recent PostsI'm not sure where you have read that the Neumann boundary should be in the range 1 to 100? In section 4.5.6.1 of the manual it gives confusingly gives various different limits but for

**Timeseries Neumann**boundaries it gives the range -100 to +100.

If your river gradient is 0.01% then your Neumann boundary should be approximately 0.0001 (or -0.0001 depending on the grid orientation i.e. if the water slope is downwards in the positive M/N direction then the sign of the Neumann boundary will be negative I think.

You may want to consider starting off with a simple constant water level boundary condition to get your model running and give it an initial test, then move to a Neumann one - it's always good to start simple!

Cheers,

Richard

##### Khaled Hassan, modified 2 Years ago.

#### RE: Boundary Condition

Youngling Posts: 1 Join Date: 9/26/17 Recent PostsI'm using Delft 3D for my master thesis, I'm trying to model a flooding event in a river. The river has 2 inflows and 1 outflow, so i have three boundary conditions, for the inflow boundary conditions,I'm using Time Series Discharge flow, and for the outflow I'm using Neumann gradient of 0.09%. My problem is after I run the simulation is that, near the Neumann boundary condition at the end of my river, I have zero flow when I review the results and this doesn't add up.

The water levels there are also the same as they were before the begining of the simulation, I tried to add the Neumann number as 0.09, -0.09, 0.00009, and -0.00009, but nothing is working, it's always the same problem. ANy help or insight on how to deal with this problem will be helpful.

I have attached a zip file with the entire project in it, in case it's needed.

Thanks in advance.

### Attachments:

##### Rose Sutton, modified 5 Months ago.

#### RE: Boundary Condition

Youngling Posts: 1 Join Date: 10/12/19 Recent PostsHi Ben,

When I am building river models there are three options I consider for downstream boundaries, each with different strengths and weaknesses:

Option 1: Time varying (or constant) water level boundary

Option 2: Water level vs flow boundary

Option 3: Neumann (water level gradient) boundary

Option 1 gives you complete control of the boundary condition and is relevant if you have reliable data on downstream water level (e.g. there is a water level recorder at your downstream boundary or your river ends at a lake or the sea. It is difficult to apply if you don't know in advance what the water level should be, and it cannot cope well with water level variation across the length of the boundary (i.e. across the width of the river, five nights at freddy's free online).

Option 2 can be specified by doing a 1D normal depth calculation to generate a rating curve outside of delft3d, or by taking a rating curve from some other source (e.g. if you're boundary is at a flow gauging station with a known water level vs flow relationship). The advantage of this option is that it can cope well with time varying flows as the boundary responds to the change in flow. Disadvantages are that it can't deal with temporally varying backwater effects (e.g. in a tidal reach) and (like option1) it can't cope with water level variation across the width of the river. For stability purposes I recommend extending the rating curve well beyond the expected flow range in case there are any local flow variations (e.g. waves during model spin up from initial conditions).

Option 3 only requires you to specify a slope and can be a good solution. It has the advantages that it allows flow variability across the width of the river (important in a braided river or similar) and it allows the model to determine the water level (good if you don;t have any other data). It is important to remember that the slope you specify is the

__local__slope - so is strongly influenced by local bed roughness and bathymetry. This means you need to be very careful with roughness and bathymetry at your boundary and if you make changes (for example to roughness during calibration) then this can have a big effect on boundary water level. In the past I have found this boundary to be less stable for some rivers.

I have used all three options with success in different models, it is just a matter of selecting the most appropriate for your example.

Hope that helps,

Richard

##### Kate Lauren Neigel, modified 1 Year ago.

#### RE: Boundary Condition

Youngling Posts: 5 Join Date: 8/23/17 Recent PostsHi,

I am trying to run a model with 1 upstream flow BC, and 2 downstream waterlevel BCs. Does anybody know if Delft3D flow can be used for downstream open boundary splitting? My model runs, but when I open the output it is empty (dry).