The PMIP File Formats

Note: the PMIP file formats were originally defined in Newsletter 6. Click here if you want to learn how to find the files in the database.

The README files

Click here to see examples in the database.

A README file has been provided for each model and simulation, with a name of the form : modelid_run_README.txt
where

• modelid is the standard model name.
• run has one of the following values : 0fix, 6fix, 21fix, 0cal,
  21cal (and 6cal in the case of llnl_nh1).

The lines of a README file can be either :

• blank lines.
• lines with a keyword followed by standard information.
• comment (or additional information) lines starting with a "#".

List of the keywords and their related information :

• MODEL_ID should be identical to the modelid parameter supplied to subroutine wrtascii.
• MODEL_NAME should be identical to the modelnm parameter supplied to subroutine wrtascii.
• PMIP_RUN should be identical to the run parameter supplied to subroutine wrtascii.
• YEARS_MEAN is the number of years contributing to climatological statistics (i.e. averaging period for data) and should be identical to the lensim parameter supplied to subroutine wrtascii.
• YEARS_SPUNUP is the number of years the model was "spun up" prior to the climatological averaging period (i.e., the spin-up time).
• NB_SAMPLES is the number of samples per day contributing to the time-means. If one accumulated over each time-step, then the number of samples should equal the number of time-steps per day even though one may have only saved the accumulations a few times per day. If different variables were sampled at different intervals, this parameter should be the fewest samples per day that contributed to any of the time means. The sampling interval for each variable may be listed in the comments section of the README file.
• DIM_LONGITUDE is the model's longitude dimension (i.e. the number of longitude grid cells in the model).
• DIM_LATITUDE is your model's latitude dimension (i.e. the number of latitude grid cells in the model).
• DIM_LEVELS is your model's vertical coordinate dimension (i.e. the number of levels in the model).
• MODEL_MONTHS is a list of the number of days in each month, according to the model. It is a list of blank-separated integers.
• PMIP_CONTACT is the person to contact concerning this simulation.
• ADDRESS is the contact's address.
• PHONE is the contact's phone number.
• FAX is the contact's fax number.
• EMAIL is the contact's e-mail address.

NOTE: The following final two items must appear on comment lines (i.e., lines starting with '#')

• MODEL_DOCUMENTATION is followed by a list of references that document the model.
• COMMENTS is followed by any further information of interest concerning the model and the simulation.

The binary files

Click here to see examples in the database. It may also be interesting to look at the tools' links.

The binary files are in auto-documented NetCDF format and were generated from ASCII files submitted by the PMIP participants. The paragraph about text files, below, explains how to make an ASCII dump from a binary file.

Each file contains only one variable. The file names have the following format : modelid_run_variable_stat_vartype.nc
where

• modelid is the standard model name.
• run has one of the following values : 0fix, 6fix, 21fix, 0cal, 21cal (and 6cal in the case of llnl_nh1).
• variable is the name of the PMIP variable stored in the file. You can get the list of all the (theoretically) available variables on the PMIP variables' page.
• stat has one of the following values : ca, cs, cm, va,vs,vm.
• vartype has one of the following values : map, 3d, xsec.

The text files

The original version of the PMIP database provided the data files in both binary and ASCII (text) format, because the PMIP groups had to submit their files in a unified text format. The files are now available only in binary format, to reduce the size of the database (and get rid of potential consistency problems).

If needed, you can easily get a text dump from a NetCDF binary file by using ncdump (from the NetCDF package) or ncks (from the NCO package), on UNIX/Linux machines :

• ncdump -h file.nc will show the header of the data file.
• ncdump -c file.nc will show the header of the data file and the values of the coordinates.
• ncdump file.nc > file.txt will dump the whole file (header, coordinates and the variable itself) to a text file.
• ncks -H file.nc will show the values of the coordinates and of the variable.
• ncks -H -C -s"%f\n" -v variable file.nc > file.txt will dump only the variable to a text.file (with one value per line).

You should also find similar utilities on other platforms here.

Database http access

Please note that these links are only provided as shortcuts from this page to the ftp part of the base. Make sure you also read the database access page!