forked from ESCOMP/CTSM
-
Notifications
You must be signed in to change notification settings - Fork 0
/
README.CHECKLIST.new_case
40 lines (34 loc) · 2.21 KB
/
README.CHECKLIST.new_case
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
$CTSMROOT/README.CHECKLIST.new_case 03/01/2021
This is a check list of things to do when setting up a new case in order to help ensure everything is correct. There
are lots of tiny details that need to be right and it's easy to get something wrong. So the first screening to make
sure it's right is for you to carefully check through your case and make sure it's right.
The following assumes you have created a new case and are in it's case directory.
General Checklist to always do:
- Make sure CLM_ env settings are correct
(./xmlquery -p CLM)
- Make sure you are using the correct CLM_PHYSICS_VERSION
(./xmlquery -p CLM_PHYSICS_VERSION)
- Make sure you are running the appropriate overall CLM vegetation model.
The "-bgc" option of either Satellite Phenology (sp), or
Full BioGeoChemistry (bgc), or FATES (fates)
(./xmlquery -p CLM_BLDNML_OPTS)
- Also if you are running the bgc model, check to see if you should be running the prognostic crop model
(option -crop in CLM_BLDNML_OPTS)
- Make sure the LND_TUNING_MODE is correct
(./xmlquery LND_TUNING_MODE)
- For an "I compset" make sure you are running over the right forcing years
(usually ./xmlquery -p DATM_YR)
- First and align year for streams should be the start year of a historical simulation
(./xmlquery RUN_STARTDATE)
(grep stream_year_first CaseDocs/lnd_in; grep model_year_align CaseDocs/lnd_in)
- Last year for streams should be the last year you are going to run to (or beyond it)
(grep stream_year_last CaseDocs/lnd_in)
- Make sure you are starting from appropriate spunup initial conditions
(Check the run-type with: ./xmlquery RUN_TYPE)
(check finidat for a startup or hybrid simulation: grep finidat CaseDocs/lnd_in)
(check nrevsn for a branch simulation: grep nrevsn CaseDocs/lnd_in)
- Run for a month (or some short period) and go over the log files and especially the settings and files read in them.
(For an I case you especially want to look at the lnd.log and atm.log files)
Some other suggestions on things that can be done:
- Compare namelist files to an existing case if you are doing something almost the same as a previous simulation.
- Ask another collaborator to look over your case directory