This folder contains the TRACMIP simulations with the NCAR CESM model and CAM4 physics.

The simulations were contributed by Brian Rose, University at Albany
brose@albany.edu

In addition to the prescribed TRACMIP specifications, these simulations use
- 2 degree finite volume dynamical core
- 26 vertical levels
- 365-day (no-leap-year) calendar
- sea surface albedo is set to 0.1 everywhere

The output is provided in fully aggregated netCDF files (single files containing all variables and all times).
Single variables can easily be extracted using the subsetting services of this RAMADDA server, 
or using any OpenDAP-aware analysis tool (we like the xray python package for this!)

Currently the output is not 'cmorized'. This is the 'raw' output in the format produced by CESM. 
Many variable names (and possibly units) will be different from the cmor conventions.

In addition to this 'raw' output, we intend to provide standard 'cmorized' output regridded to standard pressure levels and standard variable names.
Progress on the cmorizing has been slower than anticipated. Stay tuned!

UPDATE 4/13/2016:
We discovered a bug in our original TRACMIP setup.
There was an inconsistency in the surface albedo seen by the surface and the atmosphere, and as a consequence the simulations were not energy-conserving.
We have re-run the entire TRACMIP suite. The old (corrupted) output has been moved to the inner directory 
bad_albedo/
The files in the top level CAM4 directory are the up-to-date, energy-conserving simulations.

If you have used the CAM4 output for any analysis prior to April 13 2016, we strongly urge you to re-run your analysis with the corrected runs.
Sorry for any inconvenience.

The note below is now no longer relevant.


UPDATE 1/27/2016
We discovered an error in our initial LandOrbit run -- the longitude of perihelion was off by 180 degrees, 
so that NH solstice was occurring at perihelion rather than aphelion. 
01/27/2016: this run was RENAMED to LandOrbit_rev
A new, corrected LandOrbit run was also uploaded.


