How to align transient CO₂ forcing with shifted climate forcing in standalone CLM5

[Arup]

Hello experts!

I am using CLM5 in standalone mode and would like to confirm how the transient atmospheric CO₂ forcing is selected and aligned in DATM.

I am conducting an experiment in which the model simulation period is 1907–1960, but I want both the climate forcing and atmospheric CO₂ concentration to correspond to 1961–2014. For that, I have made the following changes:
DATM_CLMNCEP_YR_START=1961
DATM_CLMNCEP_YR_ALIGN=1907
DATM_CLMNCEP_YR_END=2014
(user_nl_datm file)
strm_year_start = 1961
strm_year_align = 1907
strm_year_end = 2010

After building the case my datm_in shows "datm.streams.txt.co2tseries.20tr 1850 1850 2014".


In particular, with the generated setting

datm.streams.txt.co2tseries.20tr 1850 1850 2014

will model year 1907 receive CO₂ corresponding to 1907 or 1961, or is there another mechanism by which DATM maps the CO₂ stream to the model years?

My objective is specifically to have:
Model year 1907 → CO₂ year 1961...
Model year 1960 → CO₂ year 2014
while keeping the model simulation calendar as 1907–1960.

If the CO₂ stream is independently controlled, what is the recommended way in CLM5 to achieve this temporal alignment?

For example, should I modify the CO₂ stream-specific settings using something like:

year_first = 1961
year_last = 2014
year_align = 1907
taxmode = extend
If this is the correct approach, should these settings be specified through user_nl_datm or by creating a custom CO₂ stream file?

I would also appreciate confirmation on whether using

DATM_CO2_TSERIES = 20tr
is appropriate for this purpose, or whether

DATM_CO2_TSERIES = 20tr.latbnd
would be preferable. My understanding is that 20tr.latbnd concerns the spatial representation of the CO₂ forcing rather than its temporal alignment, but I would appreciate confirmation.

Thank you very much for your help!

 

[oleson]

You'd need to add this to your user_nl_datm:

streams = "datm.streams.txt.CLMGSWP3v1.Solar 1907 1961 2014",
"datm.streams.txt.CLMGSWP3v1.Precip 1907 1961 2014",
"datm.streams.txt.CLMGSWP3v1.TPQW 1907 1961 2014",
"datm.streams.txt.presaero.trans_1850-2000 1849 1849 2014",
"datm.streams.txt.topo.observed 1 1 1",
"datm.streams.txt.co2tseries.20tr 1907 1961 2014"

One you start running, check your atm log file to make sure it is reading in the desired time slice of the CO2 data.
20tr is appropriate for this purpose.

 

[Arup]

Thanks, @oleson, for clearing up the problem.

However, I would like to understand how the CO₂ time series is aligned with the model simulation years.

In a normal transient simulation, is the CO₂ concentration automatically aligned with the model simulation year? For example, if I run CLM5 for 1907–1960 and want to use the CO₂ time series for the same period (1907–1960), will the model automatically use the 1907 CO₂ concentration in 1907, 1908 CO₂ concentration in 1908, and so on?

Or do I also need to explicitly set the CO₂ stream as:

datm.streams.txt.co2tseries.20tr 1907 1907 1960

to ensure that the CO₂ forcing corresponds to the 1907–1960 simulation period?

In other words, I would like to understand whether the CO₂ stream is automatically synchronized with the model simulation years, or whether its temporal alignment must always be explicitly specified through the stream configuration.

 

[oleson]

The default setting for that configuration is:

"datm.streams.txt.co2tseries.20tr 1850 1850 2014"

This means that model year 1850 lines up with the year 1850 in the data file. If you start the model at 1850, it will use the 1850 data. If you start the model at 1907 it will use the 1907 data, so yes the data file is synchronized with the model simulation years given that default setting.

 

[Arup]

Thank you, @oleson , for the confirmation.