Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

How to manipulate the restart file? (u,v) field in the upper troposphere

Dear All:I was wondering if anyone knows how to manipulate the (u,v) field in the atmosphere restart files.I made a B_2000_CAM5 run on SE ne30 grid for a few years. Then I want to play with the (u,v) field at, say 200mb, or any level that does not have orography. In the history output, I can easily find the (u,v) field (48602 grid points) at any level and the Lat/Lon of any grid point, multiply the (u,v) by 1.1 for instance, and put them back to the history output. However, when I ncdump my restart files, I was surprised to see there are a lot more grid points for (u,v) field (86400 grid points), and they don't come with Lat/Lon.If I want to modify the (u,v) field in the restart files in the way that I can see some wind burst in some region, how should I do it?Thank you. Any insight appreciated. Best Regards,Wang
 

jedwards

CSEG and Liaisons
Staff member
The SE grid is a cubed sphere and writes it's output on the cubed sphere grid, this is represented in netcdf as a one dimensional field at point lat(i), lon(i) for i=1,86400.You can use tools like ncl or nco to find and modify your region of interest.   
 
Thank you very much.When I looked at the history file, I could see the lat(i) and lon(i), for i=1,2,...,48602.When I looked at the restart file, I could not find the lat(i) and lon(i) anywhere. Does ncl or nco use some tool to find the lat/lon in the restart file?Thanks again. 
 

olson

Member
   Hi, Wang,

     On restarts, the u and v are on the SE "dynamics" grid.  The
 number of points on the SE dynamics grid (86400) is roughly
 a factor of 16/9 the number of physics gridpoints(48602).  The
 extra points on the dynamics grid turn out to just be duplicates
 of some of the physics points.  As you discovered, the lats/lons
 for the dynamics grid are not dumped to the restart file.  I have
 a fortran file on /glade modified to dump those lats/lons to the
 restart file:

   /glade/scratch/olson/modified_restart

 restart_dynamics.F90 lives in the directory components/cam/src/dynamics/se

  This copy came originally from the tag cesm2_0_beta03.  I have both
 the original and the modified copy for comparison on how to modify your
 copy.  Once modified, run the model for a timestep to get a restart
 file that will also include the lats/lons for the dynamics grid (ncol_d)

    Good luck
 
Top