Jérémie Roques International Volunteer; Hydrological mapping and model implementation

Hi;

Thanks for those hints Mathieu; this helped me understand better how Mike identify internal and open boundaries.

Best

JR

Hello everybody.

I would like to manually define open boundaries in my M21 model. Basically, those boundaries are single points boundaries.
Each time I try a simulation I get the following message : 'Single point boundary not well defined' or 'Internal boundary not well  defined'.

I have absolutely no idea were is the problem from; could someone give me some hints to solve this issue ?

Might this be linked to the land value ? Or the fact that one on my boundary is in the middle of my DEM (ie. On the coast; beyond the coastal boundary are blank values; deleted on the dfs2 file) ?

Thanks

J.R.

Hi Mathieu;

Thanks again =)
Splitting meanders into smaller polygons was indeed the best solution.
Actually I am using the 2008 version so, once again, you are right !
Wish you the best.

Jérémie

Thanks Mathieu =)

Okey; this worked indeed great but I suppose there is other constrains for the quadrangular mesh generation.

Actually, the “ValidatePolygon() return FALSE” error message appear to be frequent even if the symmetry is respected regarding to the amount of points. Indeed, I managed to draw large polygons as a first test (high distance between river banks and sides of the polygon) but when I try to refine polygons (by adding points and reduce the distance between banks and polygons sides) for quadrangular mesh, the message appear and mesh generation is not anymore possible. I don’t have any idea where the message is from; any links with meanders (which are numerous) ?

The “Transfinite interpolation" option seems to not be accessible in my version (not possible to select it…).

Thanks again. Best

Jérémie

Hello everybody;

I have an issue trying to work on a Mike Zero mesh generation file.

I work on a large floodplain and would like to get better mesh resolution along river branch and lower resolution on floodplain. This is to find the best compromise between computing time and results quality.

I managed to apply quadrangular mesh on a test file and have been convinced by the method. Unfortunately, I would like to apply it to the entire area and get some problems to do so.

The river system is quite complex. Attach is a schema of the network and loops (not all clockwise direction …) I try to digitalize with land/water boundaries to separate quadrangular from triangular mesh and open boundaries for upstream/downstream conditions. There is a total of 6 portions to be mesh with quadrangular cells.

When polygon is selected; Start and End arcs defined; I basically get this message: “The polygon subdivision fail due to the unequal points on both side”; then “ValidatePolygon() return FALSE”

Does somebody could help me to solve my problem and to remesh the DEM? Any other advice for quadrangular mesh generation?

Thanks

J.R

Hi everybody

I would like to model the effect of lake within a large natural  floodplain (1000 km2) with multiple inflow/outflows.

I would ask for orientation on options I have since I am novice in flood modeling and that I do not really know which function I better have to use: Regulating or user defined structure.

Basically, maps are expected as outputs. How to be sure extends of the reservoir would be well represented? By digging the DEM (30m) or creating new branches with large interpolated CXs ?

Thanks a lot

Jérémie

Hello;

Thanks for this answer. Problem solved !
Best

Jérémie

Hello everyone;

I have an issue related to boundary I Mike 21
• Upstream boundary is file type 0 : Inflow
• Downstream boundary is file type 0 : water level (tide)

Model area is around 1000 km2. The model (inland) is surrounded by land value
Boundaries are represented by one point each within the land value belt. They are automatically detected by the software.
The river network is clearly defined (dug) within the bathymetry.

Problem is that boundaries does not have effect and are not considered within the (incomplete) result file (only the initial surface elevation does).

I tried different FAB type; no differences.

Does somebody have an idea of the origin of my problem?
Thanks by advance.

J.R

Hello;

Thanks Mathieu for this new answer, this make now complete sense when I look free space on my disk…
Well; I am using a rectangular grid and the courant number was the estimated one of the simulator. So enlarging grid size might be the only solution I have…
I will try this with a mesh grid as well.
Thanks a lot
Best

Jérémie

Hi all;

By running my Mike 11 with flood mapping option and mike 21 models I respectively get following messages:
- “error. DfsWriteItem TimeStep return code is 2002”
- “WriteDataItem: Write error"

Does those messages mean that my computer miss memory?

Actually; I wanted to reduce Courant Number and tried to switch from a 1h to a 15m time step (CoN from 634 to 158; still too much… ).
Is there a better way to reduce CoN; enlarging channels?

Thanks

Jérémie