CESM Climate Variability Working Group: FAQs

Q1: I do not have access to the mass store or SCD machines, how should I get the data? Can I ask the CVWG liason to get it for me?

  • The CVWG currently doesn't have a liason. For small requests of post-processed data, you can email me here, and I will be happy to copy the data to CGD's ftp site. For all other requests it is suggested that you get an SCD account, you can get that by clicking here.

Q2: I noticed that the runs listed as CAM3 runs used a version of cam called cam2_0_2, and that in the history file names "cam2" is present, are these runs really CAM3 runs?

  • Yes, cam2.0.2 is officially CAM3. Recent versions of CAM3 have been named cam3.1.* to avoid this type of confusion. Technically, all versions of cam2.0.2 -> cam3.0 are referred to as CAM3.0, while all versions of cam3.1 are referred to as as CAM3.1. CAM3.1 reproduces the same climate as CAM3.0. Please visit the CAM webpage for more details.

Q3: Some of the post-processed data is in 20-year chunks per netCDF file, how can I merge all these 20-year files together into one file?

  • It is suggested that you use the netCDF operators to concatenate the files together. The operators are available on most CGD/SCD systems. In this case you would use ncrcat, which will concatenate a number of files into one file:
    ncrcat b30.009.cam2.h0.Z3.0400-01_cat_0419-12.nc b30.009.cam2.h0.Z3.0420-01_cat_0439-12.nc/ptmp/asphilli/b30.009.cam2.h0.Z3.0400-0439.nc

Q4a: When I do a ncdump on a CAM3 monthly history file, the date inside the file doesn't match the file name, why?
Q4b: When I read a CAM3 file into GrADS, GrADS says that the time is a different time from the file name. Which is correct?

  • For monthly history data, the name of the file is always correct. The time in the file (denoted by the variable date) is the model time that the previous monthly average was calculated. Thus, if the file name is newsstamip31.cam2.h0.1950-01.nc, and the date variable inside the file is set = 19500201, that file contains the January 1950 monthly average (which was calculated on February 1st, 1950).
  • For non-monthly history files (ex. daily), let's use the file eul128x256_d63vamip_trpsst_04.cam2.h1.2001-03-15-00000.nc as an example. The first timestep would contain the March 14th daily average for averaged variables, it would contain the March 15th 00Z instantaneous value for instantaneous variables. This can be deduced by looking at the variables time, time_bnds, date, and datesec.

Q5: Is there a reference for the SST dataset that is used for the CAM2/CAM3 AGCM runs?

  • Hurrell, J.W., J.J. Hack, D. Shea, J.M. Caron, and J. Rosinski, 2008: A new sea surface temperature and sea ice boundary data set for the Community Atmosphere Model Journal of Climate: submitted as a note.
  • The SST dataset is a blended version of the HadISST and Reynolds datasets.

Q6: What programming language is recommended for processing/visualizing CCSM/CAM model data?

  • It is recommended that you use NCL. NCL is easy to use, can read in many different types of data (netCDF, HDF4, HDF4-EOS, GRIB, binary and ASCII), runs on most system architectures (Solaris, AIX, IRIX, Linux, MacOSX, Dec Alpha, and Cygwin/X running on Windows), has a regularly updated web page with many example scripts, produces high quality graphics, has a support team, and best of all is free. For a more thorough overview please click here.