Previous Section Headers
3.5.1 History Files
History files contain descriptive information and data values written
at specified times during the run. As discussed in
Sec. 2.5 , the model automatically
writes a "primary" history file, and has the capability of writing up
to five "auxiliary" history files if namelist variable
INITHIST is set to 'NONE' and up to four otherwise.
A "Master Field List" ( Table 2.13
) contains a list of all the variables that can be output to a
history file. Certain fields on the Master Field List appear by
default on the primary history file (as indicated in
Table 2.13 ). Using the CCMEXP
namelist variables PRIMARY , EXCLUDE , and
AUXF , the contents of each history file can be customized.
Other namelist variables control the time frequency at which each
history file is written, along with the type of field averaging
(averaged, instantaneous, minimum, maximum, or monthly-averaged),
packing density, and number of time samples written to a single
history file. The NCAR Mass Store System can be used to archive
history files. History files are written directly in native machine
binary format rather than netCDF due to performance considerations and lack
of packing support. However, we support a conversion tool
(ccm2nc)
that converts CCM history files into netCDF. In general any program
that reads in a CCM history file, will need to be run on a machine that
stores data in the same format as the machine that created the CCM history
file. But, ccm2nc is able to recognize and convert IEEE or Cray COS
blocked formats from either class of machines. However, currently ccm2nc
is not able to recognize or convert byte-swapped data-files.
Sub Sections
-
3.5.1.1 History File Format
-
-
3.5.1.2 History File Header Records
-
-
3.5.1.3 Latitude Data Records
-
-
3.5.1.4 History Buffer
-
-
3.5.1.5 History File Code Flow
-
Search page
Questions on these pages can be sent to...
erik@ucar.edu .
$Name: ccm3_6_6_latest2 $ $Revision: 1.38.2.1 $ $Date: 1999/03/25 21:38:41 $ $Author: erik $