Spinup stability test in CLM5 (CESM2.1.3) ND case, peculiar issue!

[wvsi3w]

Hello dear Keith @oleson and Sam @slevis
This image is the output of Spinupstability tool for a case of mine that was running smooth but suddenly I see an increase in all variables (around year 1850 which is 850 for ND case) and is very odd and strange:


Do you know by any chance what would be the reason for this?
I havent seen this before.

Should I remove the outputs from 1850 onward and redo the test?

P.S. The reason why I continued the spin-up for this ND case was to decrease the "% of land area in TOTECOSYSC Disequl" to be below 3 percent.

 

[oleson]

The fact that this happens around 1850 is suspicious. I wonder if the model has transitioned to using transient (which usually starts in 1850) data instead of constant data somehow. There was a bug associated with carbon isotopes if you ran past 1850 years in a spinup (Use of C13/C14 timeseries for non-transient control cases gives incorrect input when run long enough · Issue #592 · ESCOMP/CTSM), but I didn't think it affected anything else in the model. Typically, we start model spinups at year 1 since they aren't "real" years. In other words, 1850 is just the number of years, not the actual year.
I would look at your log files (mainly atm and lnd) around the 1850 spinup year to see if the forcings have changed somehow, e.g., these will likely be from the streams files, e.g., ndep.

 

[wvsi3w]

The fact that this happens around 1850 is suspicious. I wonder if the model has transitioned to using transient (which usually starts in 1850) data instead of constant data somehow. There was a bug associated with carbon isotopes if you ran past 1850 years in a spinup (Use of C13/C14 timeseries for non-transient control cases gives incorrect input when run long enough · Issue #592 · ESCOMP/CTSM), but I didn't think it affected anything else in the model. Typically, we start model spinups at year 1 since they aren't "real" years. In other words, 1850 is just the number of years, not the actual year.
I would look at your log files (mainly atm and lnd) around the 1850 spinup year to see if the forcings have changed somehow, e.g., these will likely be from the streams files, e.g., ndep.

Thank you Keith for your response.
I agree. I checked my case (logs directory + run directory + lnd directory + everywhere) and I couldn't find the log file for land and atmosphere. this is odd. however the stream files for the run directory is showing 1901-1902 forcing (the years I assigned). I will try to find some related log files, I don't know why I don't have them.

Also, I realized that around year 1750 the percentage of that disequilibrium of land area soil carbon reached 3% (I moved anything above 1750 to another directory and ran the spinupstability.ncl). So I think my spin-up is now done for this case with 750 years of ND decomposition.

Thanks.