Tidal mixing in a deep-time paleo simulation

[Siva]

Hello. Requesting help with my query.

I am running a mid-cretaceous simulation with CESM1.2.2 and I prepared the setup using the paleoclimate toolkit. I had to disable the tidal mixing component as mentioned in the document. Otherwise, it gives the pop convergence error and reducing the pop2 timestep even by a large margin doesn't solve the problem.

I am interested in the ocean circulation and mixing analytics of this period. So my question is, is there a way I enable the tidal mixing(the ltidal_mixing flag) and run my deep-paleo simulation? Does preparing a period-specific tidal forcing help to solve the issue and if so, are there rules to be followed for preparing the forcing to avoid pop2 errors like the convergence error?

Thank you

 

[Siva]

Dear all,

The issue is not resolved yet so any suggestion would be very helpful.

To make it more clear, I am trying to run my deep-time simulation (tried both startup and hybrid) with the tidal mixing component turned on. (keeping lhoriz_varying_bckgrnd, ldiag_velocity, ltidal_mixing to TRUE while keeping the overflows tags OFF). I prepared the tidal energy nc file to the same dipole grid and the land-sea mask as that of my kmt file. Also added the necessary dimensions and variables like TAREA to the nc file. But it doesn't work and gives the "POP aborting... Error in getting varid for netCDF field" error. Additionally, turning on the overflows tags gives the "kmt inconsistency for overflows" error. I just want to know if I am doing it the right way or missing a step somewhere. Thank you.

 

[macarew]

Hello Siva,

I have never tried to run a deep-time simulation with tidal mixing turned on (always 'ltidal_mixing = .false.'). However, I have some ideas that may help you debug your simulation (hopefully this is somewhat helpful): (1) Try to get the case running with tidal mixing off first to make sure the errors are coming from the tidal mixing file when you set 'ltidal_mixing = .true.'. If you are still getting the 'POP convergence error' and have already tried increasing dt_count, then there may be another issue with your TS initial condition (init_ts_file) to focus on first. (2) Have you taken a look at the default modern tidal mixing netCDF files on Cheyenne? For example see tidal_energy files here: /glade/p/cesmdata/cseg/inputdata/ocn/pop/gx1v6/forcing/. This may help confirm that you are replicating the exact netCDF format for your POP resolution.

 

[Siva]

Hello Sophia,

Thank you very much for your solution. It gives me the idea that it is possible to run a deep-time simulation with the mixing turned on but with the right setup. I should have mentioned that the simulation works fine with the ltidal_mixing = .false. So as you suggested, the problem could be with the tidal mixing netCDF file I prepared. I will check it with the sample files from the repository and will update if it works.

 

[LisaCickyham]

Thanks for your post

 

[JamieFernandez]

Dear all,

The issue is not resolved yet so any suggestion would be very helpful.

To make it more clear, I am trying to run my deep-time simulation (tried both startup and hybrid) with the tidal mixing component turned on.

(keeping lhoriz_varying_bckgrnd, ldiag_velocity, ltidal_mixing to TRUE while keeping the overflows tags OFF). I prepared the tidal energy nc file to the same dipole grid and the land-sea mask as that of my kmt file. Also added the necessary dimensions and variables like TAREA to the nc file. But it doesn't work and gives the "POP aborting... Error in getting varid for netCDF field" error. Additionally, turning on the overflows tags gives the "kmt inconsistency for overflows" error. I just want to know if I am doing it the right way or missing a step somewhere. Thank you.

Download the best and free animation apps at Techzapk. Techzapk provides a list of the best animation viewing apps from various sources and developers. This site provides detailed information about each app, including reviews, features, user interface, and animations.

It seems like you are experiencing some issues with running your deep-time simulation with the tidal mixing component turned on. I understand that you have prepared the tidal energy nc file and added the necessary dimensions and variables. However, you are encountering an error related to getting the varid for the netCDF field, and enabling the overflows tags results in a "kmt inconsistency for overflows" error. To fix this problem, there are several steps you can take from me: Double-check the compatibility: Ensure that the dimensions, variables, and grid configurations in your tidal energy nc file match the requirements of your deep-time simulation. Pay attention to factors like grid resolution, land-sea mask, and any specific formatting or naming conventions. Verify variable names and attributes: Check that the variable names and attributes in your tidal energy nc file are correct and consistent with what the model expects. Pay attention to case sensitivity and spelling. Review input file settings: Check the configuration settings in your input files, such as the namelist options and runtime options related to the tidal mixing component. Ensure that the options are correctly specified and aligned with your intended simulation setup. Consult model documentation or user community: Refer to the documentation or user forums specific to the model or software you are using for your deep-time simulation. These resources may provide insights into common issues, troubleshooting steps, or specific requirements for incorporating tidal mixing. Seek assistance from experts: If the issue persists, consider reaching out to experts or the model developers for further guidance. They may be able to provide specific recommendations based on your model setup and the errors you encountered. Remember to provide specific details about your model version, input files, and the exact error messages you are encountering when seeking assistance, as this will help others better understand and troubleshoot the issue. I hope these suggestions are helpful, and I wish you success in resolving the issue with your deep-time simulation.

 

[LisaCickyham]

Hello. Requesting help with my query.


I am running a mid-cretaceous simulation with CESM1.2.2 and I prepared the setup using the paleoclimate toolkit. I had to disable the tidal mixing component as mentioned in the document. Otherwise, it gives the pop convergence error and reducing the pop2 timestep even by a large margin doesn't solve the problem.


I am interested in the ocean circulation and mixing analytics of this period. So my question is, is there a way I enable the tidal mixing(the ltidal_mixing flag) and run my deep-paleo simulation? Does preparing a period-specific tidal forcing help to solve the issue and if so, are there rules to be followed for preparing the forcing to avoid pop2 errors like the convergence error?

Thank you

Visit Techzapk to discover free and safe technology apps.

Hello, I understand your query regarding enabling tidal mixing for your mid-cretaceous simulation with CESM1.2.2. While disabling tidal mixing can help avoid convergence errors, you're interested in including it for ocean circulation and mixing analytics during that period. To enable tidal mixing, you can try preparing a period-specific tidal forcing dataset that corresponds to the mid-cretaceous era. This may help resolve the convergence issues you faced previously. When preparing the forcing dataset, it's crucial to ensure it's compatible with the model and follows the appropriate guidelines to avoid pop2 errors. Review the CESM1.2.2 documentation or seek advice from experienced users for specific rules and best practices on preparing tidal forcing to prevent convergence errors. Wishing you success in your deep-paleo simulation and ocean analytics! Thank you.

 

[Siva]

Hello all. Thank you for the detailed replies and guidance. I was able to solve this issue a few weeks ago by following several of your suggestions. To summarise the steps I followed to prepare a tidal energy file in the pop grid format and run a deep time simulation with tidal mixing turned on,

• I used a modified version of the ESMF_regrid_28.ncl script for regridding the tidal energy file from regular to pop grid. I used a pop history file from my earlier model output as the target grid for a bilinear interpolation.
• Using the same script, I added the variables TAREA, TLAT, TLONG in the output file and had to check for consistency with the CESM tidal mixing file on Cheyenne. The Dtype of the final output file has to be F64 encoding or the file didn’t work in the model.
• With this file, the model works with tidal mixing on ('ltidal_mixing = .true.') But the ‘overflows’ options had to be turned off. The model setup with significantly different deep-time geography runs only with higher values of dt_count.

Thanks again,
Siva

 

[macarew]

Nice work @Siva ! Thank you for reporting your process/solution here.