The initial condition files that PTS_MODE uses

[Xueli Huo]

Hi

I am reading the CLM5 Documentation
1.6.2. Running a single point using global data - PTS_MODE
1.6.2. Running a single point using global data - PTS_MODE — ctsm release-clm5.0 documentation
I am confused with the sentence "PTS_MODE currently does NOT restart nor is it able to startup from global initial condition files."

My questions are
1) Why PTS_MODE currently does NOT restart ? Because the mode doesn't generate restart file ?
2) What is the initial condition file for the mode to run if it doesn't use the global initial condition file ?

Sincerely,

 

[erik]

PTS_MODE reads in global datasets and then only operates over a single point. the way it works is to have modifications throughout the code to handle global read of 2D data -- but then only run over a single gridcell. So some of the time it recognizes that the grid is a large global one, and some of the time it recognizes it's really only running a single point. As such it's not a completely robust implementation, and has some fragility.

1.) Why doesn't PTS_MODE restart?

One point of fragility is that it can't really startup from a restart file that it writes. So it really shouldn't write restart files. But, even if you set it up so that it did -- it wouldn't be able to do anything with the data it had just read in. Remember it reads in 2D datasets but then only returns a single grid point. So it has a mechanism to read in 2D datasets, but restart files are 1D vector datasets, and it doesn't have a mechanism to read them or operate on them. It would have to read in the 1D vector and then figure out the closest point to use from that. The PTS_MODE code is only setup to handle 2D datasets and not 1D vector datasets which is what CLM restart/IC (Initial condition) files are.

2.) What does it use then for IC?

PTS_MODE always does a "cold start" when starting up. The soil Temperature profile is isothermal at the surface temperature. Soil moisture is set to an arbitrary value between dry and moist. Many variables start at zero if that is a reasonable starting point. How bad it is to do a cold start depends on both what mode you are running in, and what what you are looking at as well as where on the global you are running. The arctic takes longer to spinup the soil Carbon for example. And the full biogeochemical model (BGC mode) takes longer to spinup than the Satellite Phenology mode.

 

[Xueli Huo]

Hi Erik,

Thanks for the detailed explanation.
Another question is the IC for PTS_MODE is automatically generated and is identical no matter which model mode I am running, which variable I am looking at as well as no matter which the single point that I am running ?

 

[erik]

It's automatically generated. It doesn't matter too much on the model mode you are running.

It does matter what variable you are looking at. As I said many are zero, some are set to isothermal based on the surface temperature. What the surface temperature is does depend on what data you are getting from your atmosphere model, as well as the specific point. The specific point will also change the soil characteristics and as such change the amount of soil water. So some things depend on where on the globe the specific point is.

 

[Xueli Huo]

Ok. Get it. Thanks !