stearm of docn_ocn_in

[njujyq@163_com]

I want running my simulations with the sst of year 2003.

The SST will cycle for all model years.

So I choose the SST data: sst_HadOIBl_bc_0.9x1.25_1850_2012_c130411.nc.

And I write the stream like : docn.streams.txt.prescribed 1 2003 2003.

The taxmode is cycle.

The I run CESM from year 0001-01-01.

I don't whether this setings is right.

I when I read the logs of ocean.

I find:

(shr_dmodel_readLBUB) reading file: /glade/p/cesmdata/cseg/inputdata/atm/cam/sst/sst_HadOIBl_bc_0.9x1.25_1850_2012_c130411.nc 1848
(shr_dmodel_readLBUB) reading file: /glade/p/cesmdata/cseg/inputdata/atm/cam/sst/sst_HadOIBl_bc_0.9x1.25_1850_2012_c130411.nc 1837

The model start with the time step 1848 of the sst file.

But I think it should be 1836 (1850+1836/12 = 2003).

I don't know why the model start with a large time step (1848) then a small one (1837).

I really need your help.

Best!

 

[eaton]

Using ncks to look at the time samples in this sst dataset I find:time[1836]=55860.5 date[1836]=20030116
...
time[1847]=56194.5 date[1847]=20031216

These are 0-based indices, so in Fortran style indices this correspond to time samples 1837, ..., 1848 as the 12 monthly samples for 2003.  The reason that sample 1848 is listed as the lower bound is that in the interpolation to 20030101 the December sample (1848) is the lower bound and the January sample (1837) is the upper bound when cycling on 2003. 

 

[njujyq@163_com]

I understand. Thanks for help!