Hello,
Thanks to your helpful resources, I was able to generate a fully connected grid for the Mekong Delta region using Grid+.
I am now attempting to simulate salinity using this grid and my input data in EFDC+ with the salinity module.
However, I have encountered repeated negative depth errors that cause the model to abort upexpectedly.
Initially, I thought the issue was related to bathymetry or the initial water depth, so I adjusted the bathymetry and assigned sufficient initial water depths.
Despite these changes, the negative depth errors persist.
I have also tried modifying the time step, following suggestions from forum posts describing similiar issues, but this has not resolved.
If it would help, i am more than willing to share my current model setup and input files for your review.
Thank you very much for your support and guidance.
Hi sangwooda1992
Did you use the updated model? were the suggestions addressed afterwards? if you run the model I sent straight I believe you will not get the negative depth related error
Sincerely,
BMHoa
Did you send me the updated model?
I haven’t received any email from you yet.
If i missed something, please let me know.
yes I sent you the model via email and some comments along with it. maybe you didn’t get it, i will send the model here
EEMS 12.2_R02_Fixed.zip (4.8 MB)
The main changes made to your original model include:
- Adjusting the dynamic time steps.
- Fixing issues related to SGZ layering and isolating errors, as well as resolving problems at the open boundary layers.
- Changing the coordinate system from WGS84 to UTM Zone 48.
Recommended/Required updates to the model:
- Use the Connection function in EEMS to properly link grid cells and ensure continuous flow.
- Remove inappropriate open boundaries that are located in the middle of the river.
- Assign complete salinity time series to all open boundaries when running salinity simulations.
- The southern open boundaries should be moved to the outermost grid cells, instead of being placed one cell inward.
- Open boundaries named South 2 to South 5 should be removed, as they are defined using only one grid cell that overlaps with the BC-South boundary, making them unnecessary.
Sincerely,
BMHoa
Thank you very much for helping me with my model last time.
It worked well after your corrections, and following Minhhoa’s feedback, I connected the previously unlinked grids and removed inappropriate open boundaries.
Unfortunately, the simulated salinity turned out to be much lower than the observed values, which suggests that the salinity intrusion was not sufficiently reproduced.
To address this, I reduced the grids in the estuary area by about half and redefined the open boundary conditions before re-running the model.
https://drive.google.com/file/d/1mjdcx_x7fwUTTlZI92l4e6boeFvP_-CE/view?usp=drive_link
However, although I only modified the simulation period and the grid, the model stopped running again due to negative depth values.
I would be very grateful if you could review my model (attached here) and provide guidance on the following points:
- Please help me adjust my model settings so that the simulation can run successfully.
- Could you explain the modifications you made and advise what I can do when such errors occur again?
- Do you think reducing the estuary area grids could help improve salinity intrusion inland?
- Currently, I am using a time step of 0.05. However, I believe a 1-hour interval might be sufficient. Could you let me know if applying a 1-hour time step is feasible for this model?
Thank you very much for your kind support.
1- Use the “Flatten Two Adjacent Interior Cells” function to set up grid cells at the open boundary with equal number of layers.
2- The open boundaries are not fully assigned with salinity data series, resulting in no salinity source from these open boundaries
3- The N-S; E-W connections are not fully configured
I am writing to thank you for your previous help.
I was able to successfully run the modified model in EEMS 12 after applying the three pieces of feedback you provided.
As I plan to carry out long-term simulations, I intend to run EFDC+ in a Linux HPC environment to reduce computation time. Currently, I only have access to the version of efdc.x dated 2021-09-27 (MPI), which has been difficult to use because it does not properly read the EFDC.INP file.
If possible, could you kindly provide me with the latest version of the efdc.x executable?
I sincerely appreciate your support and guidance.
