SST Daily Forcing

[yongjia]

Dear all,

It is the first time to run this CESM model (v1.2.0), and I just want to simulate the case about the 2020 summer SST. But I found that the forcing data, the temporal resoulation is monthly, and I just want to replace it with daily observed SST. So what should I do? Maybe just replace it and run without any measures to other files or settings?

Anyone who has such similar experiences?

Thank everyone

 

[erik]

Do you really want to use the older model for a more current simulation? Note, that CESM1.2.0 was released in 2013, so obviously could not have 2020 data available. But, even in asking that even our more recent model versions don't have everything updated to run for 2020. But, you also don't specify what kind of case you are running. From the context I'm guessing you want to run a "F" compset with CAM, but a prescribed data ocean, is that right?

 

[Joey Wong]

Do you really want to use the older model for a more current simulation? Note, that CESM1.2.0 was released in 2013, so obviously could not have 2020 data available. But, even in asking that even our more recent model versions don't have everything updated to run for 2020. But, you also don't specify what kind of case you are running. From the context I'm guessing you want to run a "F" compset with CAM, but a prescribed data ocean, is that right?

Hi,Erik
I want to provide daily SST forcing for FHIST. Can you tell me how should I do it? Can you give me some advice? Or Is there any tutorial available explaining the modification required in oder to run FHIST using daily SST input.
Looking foward to your reply!

 

[islas]

Hi,Erik
I want to provide daily SST forcing for FHIST. Can you tell me how should I do it? Can you give me some advice? Or Is there any tutorial available explaining the modification required in oder to run FHIST using daily SST input.
Looking foward to your reply!

Hi Joey,

I only have experience with this in CESM2, so for CESM1 I'm less confident. If you have made your own daily SST inputs, you can point the model toward reading those using the xml variable SSTICE_DATA_FILENAME and you'll need to set SSTICE_YEAR_ALIGN and SSTICE_YEAR_START and SSTICE_YEAR_END to match with your file. These are all in env_run.xml. If you set those up, there's a chance it may just work if the file is daily and automatically read in the daily SSTs and sea ice, but I'm not totally confident about this.

 

[ohmpawat]

Hi Joey,

I only have experience with this in CESM2, so for CESM1 I'm less confident. If you have made your own daily SST inputs, you can point the model toward reading those using the xml variable SSTICE_DATA_FILENAME and you'll need to set SSTICE_YEAR_ALIGN and SSTICE_YEAR_START and SSTICE_YEAR_END to match with your file. These are all in env_run.xml. If you set those up, there's a chance it may just work if the file is daily and automatically read in the daily SSTs and sea ice, but I'm not totally confident about this.

Hello, I recently need to change the SSTICE_YEAR_ALIGN, SSTICE_YEAR_START and SSTICE_YEAR_END. I know the SSTICE_YEAR_START and SSTICE_YEAR_END are the beginning and end year of my sstice data. But I don't know the meaning of SSTICE_YEAR_ALIGN. Could you tell me the meaning? Thanks!
ohmpawat

 

[islas]

Here's the info on SSTICE_YEAR_ALIGN...

The model year that corresponds to SSTICE_YEAR_START on the data file. .
Prescribed SST and ice coverage data will be aligned so that the first year of
data corresponds to SSTICE_YEAR_ALIGN in the model. For instance, if the first
year of prescribed data is the same as the first year of the model run, this
should be set to the year given in RUN_STARTDATE.

If SSTICE_YEAR_ALIGN is later than the model's starting year, or if the model is
run after the prescribed data ends (as determined by SSTICE_YEAR_END), the
default behavior is to assume that the data from SSTICE_YEAR_START to
SSTICE_YEAR_END cyclically repeats. This behavior is controlled by the
"taxmode" stream option; see the data model documentation for more
details.

Only used in F and A compsets.

 

[ohmpawat]

Here's the info on SSTICE_YEAR_ALIGN...

The model year that corresponds to SSTICE_YEAR_START on the data file. .
Prescribed SST and ice coverage data will be aligned so that the first year of
data corresponds to SSTICE_YEAR_ALIGN in the model. For instance, if the first
year of prescribed data is the same as the first year of the model run, this
should be set to the year given in RUN_STARTDATE.

If SSTICE_YEAR_ALIGN is later than the model's starting year, or if the model is
run after the prescribed data ends (as determined by SSTICE_YEAR_END), the
default behavior is to assume that the data from SSTICE_YEAR_START to
SSTICE_YEAR_END cyclically repeats. This behavior is controlled by the
"taxmode" stream option; see the data model documentation for more
details.

Only used in F and A compsets.

I'm so sorry for my poor English. I think I can't understand the discription. If my RUN_STARDATE=1996-01-01, and SSTICE_YEAR_ALIGN=1850, SSTICE_YEAR_START=1850, which year of the data will I use?

 

[islas]

I think if you have a time series from 1850 to present and you want the SSTs in year 1996 of your simulation to correspond to year 1996 of the time series then you should do SSTICE_YEAR_ALIGN=1850. But I also always find this a bit confusing, so it's best if you give that a go and make sure that it's working as you expect.

 

[ohmpawat]

Which makes me upset is that I don't know how to check the data that I use.

 

[islas]

I would just run year 1996 of your simulation and plot time series of the SST at various locations for your simulation and compare with the time series of year 1996 of your input data. I would make some line plots with both of those on the same panel to check that they are the same.

 

[ohmpawat]

I would just run year 1996 of your simulation and plot time series of the SST at various locations for your simulation and compare with the time series of year 1996 of your input data. I would make some line plots with both of those on the same panel to check that they are the same.

Thanks for your suggestion! I will try it!

 

[Newt]

Thanks for your suggestion! I will try it!

Hi, I am also trying to use the daily SST data to run an air-lnd coupled simulation, but it didn't run successfully using the above method. It seems that something was wrong in the RTM...((shr_sys_abort) ERROR: (shr_stream_findBounds) ERROR: LVD not found, all data is after yearLast; (shr_sys_abort) WARNING: calling shr_mpi_abort() and stopping). I wonder if you have solved this technical problem?

 

[erik]

LVD means Least Valid Date. And the problem is that you are trying to run for SST data that is beyond the year you are running. This also must mean that you've told the model to NOT cycle over the SST data, but to only use SST that's within your simulation. This depends on several settings include the simulation start year, the SST settings, and the align year.

 

[Newt]

LVD means Least Valid Date. And the problem is that you are trying to run for SST data that is beyond the year you are running. This also must mean that you've told the model to NOT cycle over the SST data, but to only use SST that's within your simulation. This depends on several settings include the simulation start year, the SST settings, and the align year.

Thank you!! Yes, my SST data is daily, so there are 365 times in my SST file. I set SSTICE_YEAR_ALIGN=1; SSTICE_YEAR_START=0; SSTICE_YEAR_END=0; SSTICE_DATA_FILENAME="my file" in the env_run.xml. I also pointed the SST to my SST file in the namelist. But it sill reported the same error. I wonder if you have some suggestions about it?