nikhilgarg.gju

I am currently trying to setup MIKE 3 restart simulations. I have tested my input.m3fm file and I am getting
some surprising error from MIKE3 run. I have tested the same input file on two different machines, where
the same input file is apparently working on one machine while it gives some random error on the other
machine.

To me, it seems that the error is prompted when reading the initial conditions in restart run, and the error
string is

Fatal error: AllocateVecR: Vector already allocated
Abnormal completion

This is prompted right after the initial conditions section of the Temperature/Salinity module. I have compared the files in both of my machines, where apart from the path of the files, everything else if same.

I have attached two screen captures when the simulation is setup and when the error is prompted. I have also attached the log file for the simulation setup which runs without any error and the input file for the same, though I have not attached input data as the file size are quite big.

It would be great if you can help/ guide me with this matter.

Thanks,
Nikhil

Hi,

I am using MIKE 3 for my simulation work, where I have used MIKE 3 in open ocean. I have configured my model with a hybrid sigma and z vertical grid.

I am now trying to understand how to process the data. I have looked at the example given in the Matlab toolbox. I am a bit confused on how to proceed. I understand the workflow on how to use the data in Matlab, but there is an inherent assumption that the data can be reshaped for number of elements in the
model grid and the number of layers, however when I try to do that i do not get the same number of elements as i have put in the model grid domain.

My initial thought is that when the grid is created in the output of MIKE 3, it does not provide any information of the places where z coordinates makes the stepped mountain when it encounters land with depth. I was wondering if any one has some suggestion on how to deal with it. Also, I would like to know how to provide the vertical grid setup (more like best practice). Right now I am doing more of a hit and trial.

Hi,

I am not sure how you are going about using IronPython. I have been using Python for pretty much all my input and output from DHI MIKE. You just have to proceed in same way as using Matlab Toolbox, You will need to use the .dll files which are read by IronPython or Python for .NET toolkit. It allows you to use the DHI toolbox. I have been writing a set of scripts for the same process, though i can't share my scripts, I would be more than happy to help you with the process if you give more details how you are doing it and at what stage are you getting problems.

Extracting and converting it to any other format is simple step once you figure out how to import the arrays from dfs2 files.

Hi,

Thanks for your reply. I changed the output to .dfs2 from .dfsu and I get the output which looks somewhat logical but I am still confused as the .dfs file system works where they flatten the array, meaning that a 16*25 array will become a list of 400 values. What I am not able to understand is that how is the flattening of array done. In my model, 16 directions for wave propagation and 25 frequencies. I am trying to figure out whether it is block of 16 direction per frequency or block of 25 frequency per direction.

Currently, I have seen that it makes sense if i use block of 16 per frequency. I am not sure if i make any sense so i just attached my result.

Hope, someone can agree to what I am trying to do.

Hi,

I am trying to obtain a 2D spectral values of Energy Density from MIKE 21 SW. I have tried using .dfsu file type, but I have been unable to do so till now.

I was wondering if some one can help me in this regards. Also, I wanted to know that what will the array look like in dfsu output.

cheers