MIKE21(2011,SP7) flood application

hello, i am using MIKE21(2011,SP7) to simulate flooding in my Msc research using a hypothetical case study where some urban features arrangement are presented(see attached jpg). generally, the flooding is assumed to originate from west side with discharge pattern (0 to 10m3/s in one hr). ArcGis is used to generate the bathymetry(shown in the attachment), and MIKEtools is used to generate the dfs file. also the lanvalue is set to high value. i faced the following problems:
1. i cannot apply the boundary conditions i mentioned above(i noted in FAq something similar but it was in MIKE21FM).
2. i would like to start with dry conditions, which means that the initial water level should be less than or equal to the bathymery. but it always generates errors.

please note:
1. i will use the multi-cell solver, accordingly i cannot use source options.
2. for other parameters i use the suggested values from DHI documentation and the MIKE21 for flood notes.
3. i will use the 2 other 2D models in my research, as such it is important to consider same conditions and parameters.
4. this forum doesn't allow to attach dfs files.

attached please find one of the bathymetry file, for reference.

thank you in advance for your understanding and collaboration.
Regards,
Mostafa Shebl
UNESCO-IHE, Hydroinformatics Msc student

Hi,

I'm no expert with Mike-21, but i believe that with a time-varying water level the cells along your open boundary must always be wet (i.e. flooded). With your approach and proposed bathymetry it sounds like you want to start with a dry boundary and then add flow from one side and have it drain (somewhere?).

It may be easier to use Mike-Flood (i.e. Mike-21 coupled to Mike-11) so that you can specify a time-varying hydrograph in Mike-11. That way you can route the water in one side of your model and out on the other side (if necessary). This approach will also avoid the issue of wetting and drying of cells along your Mike-21 boundary as this is handled via Mike-Flood.

Cheers,

Andrew

Hi,

MIKE FLOOD is definitely an option, but I think that you can get some way by just using MIKE 21 alone. I would suggest that you dig trenches at either end of your grid so the boundaries are always wet and then you can specify a time varying waterlevel boundary (make sure your trenches are deep enough). Make sure that you smooth a ramp from the trenches to the dry land so there is not too much of a step in the bathymetry. This is an option if you want a water-level boundary. If you want a discharge coming in, then I would still dig a trench at the downstream end to suck up the water, but I would simply apply discharge sources onto the grid at the upstream end. This way you don't need a trench.

Thank you all for the replies, with smoothing at inlet boundary (along with a trench) and applying a spatially varied initial water levels, i got the model running.However, the model introduces some errors for steep slope as a wave fron moves before the boundary feeds in the domain (see attached).

On other hand, the multi-cell solver is not working fine with blow up errors. i use the coarse grid as 6x6 and fine grid as 2x2. i coarsened the fine grid using the resampling tool (ARCGIS9.3) with Bilinear interpolation.

I appreciate it if anyone can help

Mostafa