template.rc 6.15 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
! CarbonTracker Data Assimilation Shell (CTDAS) Copyright (C) 2017 Wouter Peters. 
! Users are recommended to contact the developers (wouter.peters@wur.nl) to receive
! updates of the code. See also: http://www.carbontracker.eu. 
!
! This program is free software: you can redistribute it and/or modify it under the
! terms of the GNU General Public License as published by the Free Software Foundation, 
! version 3. This program is distributed in the hope that it will be useful, but 
! WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS 
! FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. 
!
! You should have received a copy of the GNU General Public License along with this 
! program. If not, see <http://www.gnu.org/licenses/>. 

Peters, Wouter's avatar
Peters, Wouter committed
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
! author: Wouter Peters 
!
! This is a blueprint for an rc-file used in CTDAS. Feel free to modify it, and please go to the main webpage for further documentation.
!
! Note that rc-files have the convention that commented lines start with an exclamation mark (!), while special lines start with a hashtag (#).
!
! When running the script start_ctdas.sh, this /.rc file will be copied to your run directory, and some items will be replaced for you.
! The result will be a nearly ready-to-go rc-file for your assimilation job. The entries and their meaning are explained by the comments below.
!
!
! HISTORY:
!
! Created on August 20th, 2013 by Wouter Peters
!
!
! The time for which to start and end the data assimilation experiment in format YYYY-MM-DD HH:MM:SS
30

31
32
time.start          : 2012-01-01 00:00:00
time.finish         : 2012-02-01 00:00:00
Peters, Wouter's avatar
Peters, Wouter committed
33
34
35
36
37
38
39

! Whether to restart the CTDAS system from a previous cycle, or to start the sequence fresh. Valid entries are T/F/True/False/TRUE/FALSE

time.restart        : False

! The length of a cycle is given in days, such that the integer 7 denotes the typically used weekly cycle. Valid entries are integers > 1

40
time.cycle          : 1
Peters, Wouter's avatar
Peters, Wouter committed
41
42
43

! The number of cycles of lag to use for a smoother version of CTDAS. CarbonTracker CO2 typically uses 5 weeks of lag. Valid entries are integers > 0

44
time.nlag           : 2
Peters, Wouter's avatar
Peters, Wouter committed
45
46
47
48

! The directory under which the code, input, and output will be stored. This is the base directory for a run. The word
! '/' will be replaced through the start_ctdas.sh script by a user-specified folder name. DO NOT REPLACE

49
dir.da_run          : template
50

Peters, Wouter's avatar
Peters, Wouter committed
51
52
53
54
! The resources used to complete the data assimilation experiment. This depends on your computing platform.
! The number of cycles per job denotes how many cycles should be completed before starting a new process or job, this
! allows you to complete many cycles before resubmitting a job to the queue and having to wait again for resources.
! Valid entries are integers > 0
55

brunner's avatar
brunner committed
56
da.resources.ncycles_per_job : 1
Peters, Wouter's avatar
Peters, Wouter committed
57
58
59
60
61

! The ntasks specifies the number of threads to use for the MPI part of the code, if relevant. Note that the CTDAS code
! itself is not parallelized and the python code underlying CTDAS does not use multiple processors. The chosen observation
! operator though might use many processors, like TM5. Valid entries are integers > 0

brunner's avatar
brunner committed
62
da.resources.ntasks : 1
63

Peters, Wouter's avatar
Peters, Wouter committed
64
65
66
! This specifies the amount of wall-clock time to request for each job. Its value depends on your computing platform and might take
! any form appropriate for your system. Typically, HPC queueing systems allow you a certain number of hours of usage before 
! your job is killed, and you are expected to finalize and submit a next job before that time. Valid entries are strings.
67

Peters, Wouter's avatar
Peters, Wouter committed
68
69
70
71
72
73
74
da.resources.ntime  : 04:00:00

! The resource settings above will cause the creation of a job file in which 2 cycles will be run, and 30 threads 
! are asked for a duration of 4 hours
!
! Info on the DA system used, this depends on your application of CTDAS and might refer to for instance CO2, or CH4 optimizations.
!
75
76

da.system           : CarbonTracker
Peters, Wouter's avatar
Peters, Wouter committed
77
78
79

! The specific settings for your system are read from a separate rc-file, which points to the data directories, observations, etc

brunner's avatar
brunner committed
80
da.system.rc        : da/rc/carbontracker_random.rc
Peters, Wouter's avatar
Peters, Wouter committed
81
82
83

! This flag should probably be moved to the da.system.rc file. It denotes which type of filtering to use in the optimizer

brunner's avatar
brunner committed
84
da.system.localization : None
85

Peters, Wouter's avatar
Peters, Wouter committed
86
! Info on the observation operator to be used, these keys help to identify the settings for the transport model in this case
87
88

da.obsoperator         : TM5
Peters, Wouter's avatar
Peters, Wouter committed
89
90
91
92
93
94

!
! The TM5 transport model is controlled by an rc-file as well. The value below refers to the configuration of the TM5 model to 
! be used as observation operator in this experiment.
!

95
da.obsoperator.home    : ${HOME}/TM5
Peters, Wouter's avatar
Peters, Wouter committed
96
97
98
99
100
101
da.obsoperator.rc      : ${da.obsoperator.home}/tm5-ctdas-ei-zoom.rc

!
! The number of ensemble members used in the experiment. Valid entries are integers > 2
!

brunner's avatar
brunner committed
102
da.optimizer.nmembers  : 10
103

Peters, Wouter's avatar
Peters, Wouter committed
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
! Finally, info on the archive task, if any. Archive tasks are run after each cycle to ensure that the results of each cycle are
! preserved, even if you run on scratch space or a temporary disk. Since an experiment can take multiple weeks to complete, moving
! your results out of the way, or backing them up, is usually a good idea. Note that the tasks are commented and need to be uncommented
! to use this feature.

! The following key identifies that two archive tasks will be executed, one called 'alldata' and the other 'resultsonly'. 

!task.rsync : alldata onlyresults

! The specifics for the first task. 
! 1> Which source directories to back up. Valid entry is a list of folders separated by spaces
! 2> Which destination directory to use. Valid entries are a folder name, or server and folder name in rsync format as below
! 3> Which flags to add to the rsync command
! The settings below will result in an rsync command that looks like:
!
!       rsync -auv -e ssh ${dir.da_run} you@yourserver.com:/yourfolder/
!

!task.rsync.alldata.sourcedirs : ${dir.da_run}
!task.rsync.alldata.destinationdir : you@yourserver.com:/yourfolder/
brunner's avatar
brunner committed
124
!task.rsync.alldata.flags g -auv -e ssh
125

Peters, Wouter's avatar
Peters, Wouter committed
126
! Repeated for rsync task 2, note that we only back up the analysis and output dirs here
127

Peters, Wouter's avatar
Peters, Wouter committed
128
129
130
!task.rsync.onlyresults.sourcedirs : ${dir.da_run}/analysis ${dir.da_run}/output
!task.rsync.onlyresults.destinationdir : you@yourserver.com:/yourfolder/
!task.rsync.onlyresults.flags : -auv -e ssh
131