Plotting GCHP Output


Panoply is useful for quick and easy viewing of GCHP output. Panoply is a grahpical program for plotting geo-referenced data like GCHP’s output. It is an intuitive program and it is easy to set up.


You can read more about Panoply, including how to install it, here.

Some suggestions
  • If you can mount your cluster’s filesystem as a Network File System (NFS) on your local machine, you can install Panoply on your local machine and view your GCHP data through the NFS.

  • If your cluster supports a graphical interface, you could install Panoply (administrative priviledges not necessary, provided Java is installed) yourself.

  • Alternatively, you could install Panoply on your local machine and use scp or similar to transfer files back and forth when you want to view them.


To get rid of the missing value bands along face edges, turn interpolation off in the Array(s) tab of your plot settings.


Todo: see the stretched-grid tutorial’s plotting section in the meantime.

Some notes:

  • xarray and cartopy are the fundamental tools

  • cartopy > 0.18 fixes the “streaking” of grid-boxes crossing the antimeridian with pcolormesh(). As of writing, cartopy 0.19 is not yet released. In the meatime you can install it from GitHub with

    $ pip install git+
  • The cubed-sphere grid is a curvilinear grid, so you need grid-box corners to plot cubed-sphere data with pcolormesh(). See the stretched-grid tutorial for an example.