Preparing an SST boundary file for CAM3.1

[ianderson]

Hello,

I am trying to use the Merged Hadley OI.v2 SST and Sea Ice dataset as a boundary file for a CAM simulation for the period 1979-2006.
I have used the utilities regrid and bcgen supplied with CAM (060418 version) to create a boundary file on the 64x128 grid, and have succesfully run it for 2 months in cyclic mode. However, when I try to run with sstcyc = .false. the model complains about dates:

SSTINI: Failed to find dates bracketing ncdate, ncsec= 19781231 85200

I am using the standard initial data, and setting
start_ymd = 19790101
start_tod = 0

My settings in the bcgen namelist are:
mon1 = 1
iyr1 = 1978
monn = 12
iyrn = 2007
mon1rd = 1
iyr1rd = 1979
monnrd = 12
iyrnrd = 2006
mon1clm = 1
iyr1clm = 1982
monnclm = 12
iyrnclm = 2001
mon1out = 1
iyr1out = 1979
monnout = 12
iyrnout = 2006

Is this a correct namelist for running for the period Jan 1979 - Dec 2006?
Does bcgen do everything necessary to prepare an SST boundary file for CAM or are there other steps?
The format of the time axis in the file produced by bcgen is 'days since 1979-01-01', whereas the standard CAM SST file has dates in the format '%Y%m%d.%f'

Thanks for any help,
Iain

[Ilit]

Dear Iain,
I am trying to use the regrid and bcgen utilities and have some questions.
I am manipulating the SST and ice_cov variables and then want to insert them to the model as boundaries (from T42 to T42, so I don't need to change resolution).

I understand that I need 2 separate files: one for SST and one for ice_cov, but I am not sure that I build them correctly (I get an error: "NetCDF: Invalid dimension ID or name" ).
Do you have such files for example (maybe the files MODEL.had+oiv2.sst.mnly.49-01.unf.nc
and MODEL.had+oiv2.ice.mnly.49-01.unf.nc that are in the description)?

Many thanks,
Ilit

[ianderson]

Problem solved.
Because I want to start the model on 1st January 1979 and the monthly SST data is labeled the 16th, I need to include December 1978 on the SST file.

Hello Ilit,
I'm really a beginner with the CAM model, so perhaps someone else is better placed to help, but I can point you towards where I downloaded the Merged Hadley SST/Ice datasets. They are available from the NCAR Community Data Portal (you have to register):
http://cdp.ucar.edu/browse/browse.ht...ed.thredds.xml

Iain

[Ilit]

Thanks Iain.
I will post a new question about it.

[skywater]

Hi,Iain

I meet a problem when regrid the sst and ice data to T42 data set from 1870-2000. I used icesst tool of 060418 version.
My steps is that:

make
cd regrid
REGRID.p1 T42
./regrid -i xxxx.ice.xxx.nc -s xxxx.sst.xxx.nc -o regrid.T42.nc

After the last step, it appeared that "Attribute not found Abort (core dumped)". And there is nothing but lat and lon in file regrid.T42.nc.

Could you give me any suggestion?

Thanks a lot!

[hannay]

You need to specify the grid you want to use using the -g option:
./regrid -i ice_inputfile.nc -s sst_inputfile.nc -g grid_T42.nc -o outputfile.nc

where:
- ice_inputfile.nc is your 1x1 input netcdf ICE concentration file
- sst_inputfile.nc is your 1x1 input netcdf SST file
- grid_T42.nc is the grid you want to use:

For T42 it will be a netcdf file with
dimensions:
lat = 48 ;
lon = 96 ;
variables:
double gw(lat) ;
gw:long_name = "gauss weights" ;
double lat(lat) ;
lat:long_name = "latitude" ;
lat:units = "degrees_north" ;
double lon(lon) ;
lon:long_name = "longitude" ;
lon:units = "degrees_east" ;

- outputfile.nc is the output file

[skywater]

I get it. And problems have been resolved. Thanks you very much!

[hannay]

New icesst tools is available at:
ftp://ftp.cgd.ucar.edu/archive/icesst/

The icesst code allows to build CAM-readable SST and sea ice concentration boundary datasets at an arbitrary resolution.

The icesst tool was part of CAM3.1. The code has been updated to recognize the new datasets being provided at:
http://cdp.ucar.edu/MergedHadleyOI
This new code will be released as a part of CAM4.