Compute Leaf Temperature and Run Sensitivity Tests

[yan Chen]

Hello,

（1）I would like to output leaf temperature, but I could not find a corresponding file or variable name. Could you advise me on how to output leaf temperature in CTSM?

（2）In addition, while running simulations at NEON sites, I tried modifying relative humidity (RH) for specific days. When I set RH = 100, the output was the same as without modification. However, when I set RH = 90, 80, or 70, the results changed as expected. Does CTSM not support simulations with RH = 100?

（3）Finally, for sensitivity tests, I created multiple new cases under /glade/u/home/yanc/CTSM_current/cime/scripts by cloning from a transient simulation (./create_clone --clone). The new cases are in the same folder of the transient simulation. When I modified the inputdata (RH for specific days) in one case (I modified the data using Python and replaced the original input file with the updated version), I found that the same changes also appeared in other cases. Is there a better way to conduct sensitivity analysis so that I can modify inputdata for one case without affecting the others?

Thank you very much for your help.

 

[slevis]

Hello @yan Chen

(1) First I tried the "git grep" command in my copy of the ctsm:
"git grep -i 'leaf temp'"
and found various references in the code for the variable of interest; however, you're interested in finding this variable in the output, so I checked the User's Guide:

1. CLM5.0 User’s Guide — ctsm release-clm5.0 documentation

which has a section titled CTSM History Fields, where I see a history field TAF, canopy air temperature. This sounds close to what you asked, though you can look through the list more carefully.

(2) It's promising that you get different results when you change to RH < 100. It seems strange that nothing happens when you set the RH to 100. To troubleshoot, you could write out the model's RH when you have NOT modified it. Is it possible that the model's RH is always 100, so setting it to 100 does nothing?

(3) If I understood correctly, all the cases point to the same inputdata and you modified files in that inputdata. Another way would be to make a copy of the inputdata for each case, have each case point to its own inputdata, and modify each one as you need.